Smart Driver Training Simulation: Save Money. Prevent

Smart Driver Training Simulation Wolf Dieter Käppler Smart Driver Training Simulation Save Money. Prevent. 123 D...

Author: Wolf Dieter Käppler

39 downloads 1401 Views 2MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Smart Driver Training Simulation

Wolf Dieter Käppler

Smart Driver Training Simulation Save Money. Prevent.

123

Dr. Wolf Dieter Käppler Forschungsgesellschaft für Angewandte Naturwissenschaften e. V. (FGAN) Neuenahrer Str. 20 53343 Wachtberg Germany [email protected]

ISBN: 978-3-540-77069-5

e-ISBN: 978-3-540-77070-1

Library of Congress Control Number: 2008927363 c 2008 Springer-Verlag Berlin Heidelberg This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks. Duplication of this publication or parts thereof is permitted only under the provisions of the German Copyright Law of September 9, 1965, in its current version, and permission for use must always be obtained from Springer. Violations are liable to prosecution under the German Copyright Law. The use of general descriptive names, registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. Cover design: WMXDesign GmbH Printed on acid-free paper 9 8 7 6 5 4 3 2 1 springer.com

Abstract

At its core, being a road user means solving constant new driving tasks in constantly changing contexts; as a form of social behavior, it extends beyond motor vehicle operation. The driver’s freedom of action means that his or her attitude, behavior and motivation are given special importance. For this reason, targeted training procedures are used to improve traffic safety. In this respect, thanks to rapidly advancing technological developments, driving simulators offer interesting possible applications, and, furthermore, advantages in terms of objectification, documentation, data capture and evaluation. As there are hardly any risks or dangers, however, the use of driving simulators requires specific training concepts which are based on an analysis of tasks, activities and boundary conditions, and which allocate other training media their place in an overall training system. This manual brings together the basic principles of education and training, modeling, task description and analysis, and the pros and cons of simulation as a training method. It describes the method used to design appropriate teaching and training programs. The main components and a taxonomy of the simulator technology are presented. As an example, an interlinked driving teaching program which has been carried out is presented, with vehicles and simulators for professional drivers. This is followed by three advanced training programs which have also been tested. These simulator training courses for professional hazardous materials and package goods drivers are based on optimized simulator-specific teaching and training matter, covering an economic driving technique, an anticipatory driving technique including rare events and a frustration-resistant driving technique, i.e. self-control. The manual is rounded off by descriptions of scripts, learning strands, measurement values, questionnaires and analysis procedures to assess training success. Organizational forms, business management calculations and staff selection processes are suggested for the actual running of simulators. These are complemented by easy-tounderstand profiles and instructions for “train the trainer” courses.

v

Preface

A good decade after the temporary end of attempts to make driving simulation into an accepted, productive teaching and training technology, new possibilities and chances are on the horizon, motivated by current EU legislation. The author has been involved, in terms of technology and content, in the development of driving simulators and has tracked their progress. This book attempts to take driving simulation seriously as a technology for teaching and training, to demonstrate possible paths for future development and to promote the formation of a community as a basis for future success. The author would like to thank all the institutions, companies and universities involved for providing him with material, and for their constant willingness to discuss matters. Special thanks go to Prof. R. Bernotat and the Research Establishment for Applied Sciences (Forschungsgesellschaft für Angewandte Naturwissenschaften e.V.) in Wachtberg, Prof. H.-P. Willumeit and Berlin University of Technology, and all staff and students, for the years of factual, financial, technological and personal support. On behalf of the above I would like to thank the translator, Anne Koth. Although the masculine gender has been chosen in the text for convenience, the information applies equally to the feminine gender.

vii

Contents

1

Introduction: Demand and Reality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 The Technological Genesis of Driving Simulation . . . . . . . . . . . . . . .

1 2

2

Teaching and Training with Simulators . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Drivers, Vehicles and Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Developing a Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Training Course Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.1 Aims and Subject Matter of Training Courses . . . . . . . . . . . . 2.3.2 Training Concept . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 Description and Analysis of Activity . . . . . . . . . . . . . . . . . . . . 2.3.4 Evaluation and Effectiveness . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4 Training Media . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.1 Simulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.2 Definitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.4.3 Developing a Model, Transfer and Validity . . . . . . . . . . . . . . . 2.4.4 Driving Simulators: Setup and Requirements . . . . . . . . . . . . . 2.4.5 Typology of Driving Simulators . . . . . . . . . . . . . . . . . . . . . . . . 2.4.6 Advantages and Limits of Simulators . . . . . . . . . . . . . . . . . . .

5 5 12 14 17 18 20 22 23 24 25 27 32 37 45

3

Basic Smart Truck Driving Training Program . . . . . . . . . . . . . . . . . . . . . 3.1 Targets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.1 Driving Skills . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.2 Legality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.3 Safety Consciousness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.4 Solidarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.5 Morality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2 Analysis of CE Driver Training Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 New Training System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Extensions to Training Course . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Computer-Assisted Instruction . . . . . . . . . . . . . . . . . . . . . . . . .

49 51 52 53 53 54 54 55 61 64 65

ix

x

Contents

4

Smart Driver Training Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.1 Driving Tasks in Public Buses, Hazardous Material and Packaged Goods Transport . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.1.1 Selection of Critical Situations . . . . . . . . . . . . . . . . . . . . . . . . . 69 4.2 Aims and Concept of the Training Course . . . . . . . . . . . . . . . . . . . . . . 72 4.3 Economical Driving . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 4.3.1 Learning Strand and Route . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76 4.3.2 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 4.4 Anticipatory Driving Technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 4.4.1 Learning Strand and Implementation . . . . . . . . . . . . . . . . . . . . 81 4.5 Frustration-Resistant Driving and Self-Control . . . . . . . . . . . . . . . . . . 82 4.5.1 Learning Strand, Disruption Scenarios and Situational Events 83 4.5.2 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 4.6 Evaluation of Training Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91 4.6.1 Notes on Evaluating the Values . . . . . . . . . . . . . . . . . . . . . . . . . 98 4.7 Example Schedule and Simulators . . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 4.8 Group Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 4.8.1 Introduction to the Program of the Day . . . . . . . . . . . . . . . . . . 105 4.8.2 Sensitization for the Topic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105 4.8.3 Dangers of the Job and Improving Safety . . . . . . . . . . . . . . . . 106 4.8.4 Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 4.8.5 Stress and Stress Management . . . . . . . . . . . . . . . . . . . . . . . . . 112 4.8.6 Integration of Training Units . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 4.9 Questionnaires . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 4.9.1 Assessment of the Day’s Training . . . . . . . . . . . . . . . . . . . . . . 114 4.9.2 Assessment of the Trainer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 4.9.3 Semantic Differentials for Driving Simulator and Driving Tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 4.9.4 Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 4.9.5 Private and Professional Situation . . . . . . . . . . . . . . . . . . . . . . 119 4.9.6 Questionnaire on Attitude toward Road Traffic Safety and Driving Style . . . . . . . . . . . . . . . . . . . . . . . . 119 4.9.7 Follow-up and Training Needs . . . . . . . . . . . . . . . . . . . . . . . . . 121 4.10 Notes on Trainers’ Qualifications, Briefings and Replays . . . . . . . . . . 122

5

Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129

6

Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135

Chapter 1

Introduction: Demand and Reality

A wide range of educational and training sectors use simulations, and their technical offshoots, simulators, e.g. flight simulators to train pilots, or ship simulators to train captains. Even rail companies use railway simulators. Driving simulators are not used to the same extent. Some large-scale projects can be found internationally, e.g. in the USA or in Germany and France, and there are a range of driving schools which use small driving simulators – usually, and tellingly, technically reworked games simulators. However, there are no defining lines to say what a driving simulator is, or what one needs to do in terms of technology and content for a device to count as one. There are also no major common policy decisions. The best model for successfully creating these defining lines is the driving simulator’s greatest competitor, the car itself. A glance at the technological genesis of the automotive industry shows that major, early policy decisions contributed to its success. As early as the start of the twentieth century, the groups concerned agreed on the specifications required by a piece of technology to be called an “automobile”: A car has four wheels and a piston engine, has space for four people and their luggage, and drives as fast as possible. Its aim was also defined and accepted: The automobile should cross fairly large distances quickly and comfortably. At the time, it was known in German as a “Rennreise-Limousine” (long-distance racing sedan), as Knie (1994) described very well. In the case of simulators, the reasons no policy decisions have been made, and no defining lines drawn up, are easy to find. Unlike flying, driving a car is a highly dynamic affair. Drivers are only a few meters away from objects they pass at a comparatively high speed. From the observer’s point of view, relatively high angular velocities are reached. Furthermore, the social traffic environment, especially in towns, stands out for the exceptionally high number of objects and people, also moving. The course of the vehicle’s movement involves high accelerations and frequencies, whether during braking or in curves. One result of this is that “road testing” has become established as a teaching department at well-known universities. There are also extensive DIN and ISO standards on how to carry one out – another indication that the process has been successful, policy decisions drawn up and defining lines

W.D. Käppler, Smart Driver Training Simulation, c Springer-Verlag Berlin Heidelberg 2008

1

2

1 Introduction

set down: a typical task for standardization committees. This has not occurred in driving simulation. As well as the technological specifications, there are also boundary conditions. One characteristic of driving is the driver’s extensive freedom to act and make decisions; this is brought into play as an advertising ploy, with good reason considering the context. However, unlike flight simulation, where pilots work through preset action routines, outside critical situations, it is precisely this “freedom” which places high demands on “valid” learning strands in driving simulators. For students to get a reasonably realistic experience of driving, the learning strands (here: vocationally oriented thematic units) must create an almost infinite range and quantity of situational variables unknown to the students. On top of this, of course, comes the price. A simple driving simulator without this necessary variety of learning strands costs at least as much as the vehicle simulated. High-tech driving simulators to train hazardous materials drivers cost many times as much, even ten times as much, as the tractor trailers in question. The tractor trailer on the road, with an expert driving instructor, also means the learning strand can be varied in a way the simulator can hardly equal, even when all technological possibilities are put to use. Flight simulators do not have this problem; quite the reverse: A flight simulator lesson generally costs only one tenth of an actual flight lesson. However, the reasons listed above are not the only cause for the failure of driving simulation as a training method. A glance at the technological genesis of driving simulation shows that the wide community of engineers, psychologists, sociologists, economists, politicians, ecologists, associations, academies and universities involved have not managed to make the necessary policy decisions and create widescope definitions. There is no explanation of what driving simulators actually are, technologically: Both a multimillion-dollar high-tech device and a control unit with a steering wheel and a monitor are called driving simulators. More serious, however, are the problems which occur due to the lack of widescope definitions for the aim, use and purpose of driving simulators. Is it for demonstration, to satisfy people’s urge to play games, to replace a vehicle, for research, risk training, education, continued education or in lieu of a driving school?

1.1 The Technological Genesis of Driving Simulation The technological genesis of the automobile demonstrates the advantages of making policy decisions early. Such decisions were made within the community, despite all the competition there, and, incredibly, still firmly apply today, more than 100 years after they were first introduced. For example, Knie (1994) believes this accounts for the success of the diesel engine, which is a piston engine, the basis for which was laid down in the form of policy decisions by MAN and others as early as 1910. It also accounts, he believes, for the failure of the NSU/Wankel project (not a piston engine), which may also have been affected by a lack of industry interest or subtle sabotage.

1.1 The Technological Genesis of Driving Simulation

3

Of course, early policy decisions of this kind also have disadvantages, as we see today; however, they are essential for the success of any new technology. They are the only way the policy-making process can set in and the transition to the actual use of the technology can take place. This use must be recognized and the construction of the technology must be tested. To safeguard the preliminary work, it must be firmly embedded in an overall political strategy. A carefully assembled body of knowledge means that definitions can be created collectively. Appropriate supplementary research must be carried out for technological and academic validation. The technological structure must be set down using a formal and informal set of rules. This manual takes up this point. It provides basic principles for the policy decisions required within the “driving simulation” community, which have yet to take shape and depend on the demands of the sector. These policy decisions are overdue, especially in view of current EU laws on driver training; after all, simulators open up whole new avenues for training. The aim is to shift about 50 percent of practical training for the truck driver’s license from the road to simulation. The German Road Safety Council, Deutscher Verkehrssicherheitsrat e.V., currently supplements its safety training courses with simulator components, see Käppler (2000). German transport companies are supplementing their rail and tram driver training courses with simulators. Driving schools and other commercial providers offer special simulator-assisted training programs for professional drivers, dealing with efficient, safe driving. International teams are developing basic principles for understanding human error. This book sums up all the findings from this work on training measures and criteria to evaluate the success of training courses. In this sense, it can be seen as a basis for discussion on decisions to be made, and is not only aimed at specialists. As well as the introduction, it is divided into six chapters with the following contents: • • • •

Drivers, Vehicles and Errors Creating Models, Teaching, Training and Simulators Smart Truck Driver Education Program Smart Hazardous Goods Driver Training Program.

Chapter 2 describes the nature of the issue, including the basic principles of motor vehicle operation and human error, from the point of view of traffic safety policy, i.e. Model Creation, Teaching &, Training with Simulators with training course design, aims, concepts and the media. A taxonomy classifying driving simulators is presented, and the validity of results obtained with simulators is discussed in detail. Chapter 3 presents a concept for simulator-assisted truck driver training. Chapter 4 is another example of use in action and presents the conceptual design and contents of the Advanced Driver Course for an efficient, anticipatory driving technique by hazardous materials drivers. It includes a training schedule, questionnaires, briefings and replays, and notes on trainers’ qualifications. Chapter 5 contains Concluding Remarks and Chap. 6 the Bibliography. Text boxes explain major terms for a better understanding of the issue.

Chapter 2

Teaching and Training with Simulators

2.1 Drivers, Vehicles and Errors Traffic systems are very different human/machine systems for monitoring and controlling very different information and deployment processes. The ergonomic view of this type of complex system is user-oriented and starts out with tasks and activities. The classic understanding of motor vehicle operation sees it as a constant or discrete closed-loop system involving the driver, the vehicle, the environment, and sees traffic as moving objects by mechanical means in order to cover spatial distances. The task of driving itself has been broken down into subtasks and given a hierarchical structure (Rößger et al., 1962; Käppler & Bernotat, 1985; Johannsen, 1990), see Fig. 2.1. On the navigation level, a roadway is selected from the traffic network. On the road guidance or handling level, the lead dimensions of course and speed are adjusted with respect to the current traffic situation, taking into account the traffic rules (e.g. overtaking maneuvers). Stabilization means operating the vehicle on the street itself even in the presence of disruptions (e.g. crosswinds) and monitoring course and speed. Differences between intended and actual variables are minimized. A driver will perceive relevant information, plan any minimization at the cognition level and act. Rasmussen (1983) drew attention to differences between the task and its actual implementation. He dealt with actual activities and disassembled them into: • Skill-based action • Rule-based action • Knowledge-based action. Skill-based action takes place without deliberate attention. It is not possible to say what information it is based on. Some examples are simple driving operations such as changing gear. In the case of rule-based action, a situation is diagnosed by recognizing a combination of symptoms. Every situation is tied together with ifthen-else rules and certain actions are associated. Some examples are the application of traffic rules, overtaking or critical driving situations. Knowledge-based action W.D. Käppler, Smart Driver Training Simulation, c Springer-Verlag Berlin Heidelberg 2008

5

6

2 Teaching and Training with Simulators ENVIRONMENT System performance

TASK

DRIVER

VEHICLE

Road traffic network

Navigation

Perception

Driving performance

Traffic situation

Handling

Cognition

Controls

Driving situation

Stabilization

Action

Vehicle characteristics

Fig. 2.1 Closed-Loop System of Driver, Vehicle, Environment (Käppler, 1993c)

includes the deliberate formulation of goals plus the design, analysis and choice of action plans. One example is searching for a route in an unfamiliar area. The overlaps between the tasks in Fig. 2.1 and these action levels are fluent. Each subtask (navigation, road guidance and stabilization) can be processed at a skill-based, rule-based or knowledge-based level. For example, an unskilled driver masters the stabilization of the vehicle with deliberate attention, i.e. knowledgebased, if he is inexperienced or if vehicle handling characteristics are problematic. However, even this representation is not yet complete. Looking at all motoring activities shows that drivers, like other operators, actually do much more than conventional models suppose. They collect and judge a lot of information about different planning levels in order to plan and implement transports. This may be classified as follows: • • • • •

Determination: “What is the situation like?” Assessment and decision: “What does this mean and should I act?” Planning action: “How shall I act?” Carrying out the action: “I act.” Checking: “Was I successful?”

Alongside technical problems which are of no further interest here, a range of disruptions can occur, such as: • incomplete information for any assessment • large number of steps, apparently of equal merit

2.1 Drivers, Vehicles and Errors

7

• consequences of actions which can hardly be evaluated • little practice in current conditions. Nonetheless, even in more difficult circumstances, the driver retains the responsibility. He carries out the assessment and makes decisions, even in case of breakdowns, technological faults or accidents. As it is understood, he always has to: • act appropriately fast, and • act safely and with fault-tolerance. What does this mean in terms of road safety? Beyond simply operating the vehicle, driving in traffic also means solving tasks in constantly changing situations. The behavior of a road user must take reality into account and incorporate the driver’s experience. Traffic regulations set down standards for behavior. They provide limits and an orientation. A driver’s behavior in traffic takes place in a physical and social environment, which affects it. It is characterized by constant confrontations with situations requiring the driver to find possible, multi-dimensional solutions. Dynamic interactions take place between his convictions, attitude and his actual behavior. The concept of traffic safety therefore does not just discuss traffic as moving objects by mechanical means, in order to cover spatial distances, but as a form of social behavior. Some additional remarks. Duncan et al. (1991) showed, for example, that there are hardly any significant correlations between accident rates and driving skills. Summala (1989) showed that truck drivers’ high qualifications or traffic experts’ knowledge were not reflected in their driving more slowly. Studies such as Aschenbrenner et al. (1976), Edwards et al. (1977), Olsen (1973), Seydel & Beetz (1978), Tränkle et al. (1988), or Utzelmann (1985) and Müller (1994) showed that serious problems with driving skills only occur with certain groups of drivers, such as novice drivers or older road users with reduced ability. It becomes clear that behavior, attitude and motivation have consequences on driving knowledge and skills (e.g. Biehl, 1976; Käppler & Voß, 1988; Käppler, 1991, 1993a, b, c, and 1997a). The link between general attitude and actual behavior in concrete situations is not clear. However, it seems clear that attitude does not necessarily cause certain behaviors, but allows us to predict them. Traffic is a social and sociological problem and a form of human behavior, as well as a complex product of basic cultural conditions, institutional rules, situation-related circumstances and personality-specific factors (Büschgens, 1993).

Behavior means any physical activity, which, unlike psychological processes, can be objectively observed by third parties. This also includes processes of experiencing, thought and wants (Dorsch, 1987/1991).

8

2 Teaching and Training with Simulators

Attitude is an individual’s tendency or predisposition to evaluate an object in a certain way; it is characterized by consistency of reaction (Katz & Stotland, 1959; Lindgren, 1973; Herrmann et al., 1977). Motivational variables explain why a person acts in a specific way and with a specific intensity under certain circumstances. As well as stimulus variables, they are important behavioral determinants (Dorsch, 1987/1991). So far, so good, but people do not do everything according to plan. For example, the isolation of the drivers in their vehicles is held responsible for decreased danger perception or lack of patience. Falling back upon earlier stages of development, known as regression (Dorsch, 1987/1991), is encouraged by the fact that car drivers, unlike pilots on scheduled flights, have markedly greater freedom to make choices and act, for example in choosing their speed. This means there is more space for the variables of attitude, motivation and behavior. Yet the driver still retains the responsibility and is, after all, expected to act quickly and safely without making any errors. Motivation means assumptions about the number of reasons for acting; these assumptions activate, control and maintain behavior despite obstacles (Wittig, 1994). This is where the concept of human error comes in. It not only addresses the concrete action in the actual situation, but discovers mistakes and their causes in detailed investigations. The concept treats errors as (unwelcome) outputs of human/machine systems and a natural product of human behavior. According to Senders & Moray (1991), actions become errors when their results are assessed. Under certain conditions, the same behavior generates positive results, under others apparently negative ones. For example, driving faster means arriving earlier. In reduced visibility or rain, however, the same driving style increases the probability of accident and, perhaps even more importantly, the seriousness of the accident. This is not wanted: in this case, we speak of speeding. Yet in both cases, the driver determines the situation, judges it, makes a plan, carries it out and checks the outcome, just with different results. After all, at the start we leave it up to the driver to decide what speed is appropriate, and trust that he will choose a speed correctly and reliably, so that there are no unwanted consequences. Ultimately, our traffic system is in fact based on trusting that people’s actions are generally reliable, safe and error-free – and yet accidents still occur. Recent attempts question this view and refer to in-depth accident analyses which, over and over again, show behavior-conditioned errors and faulty ascertaining, judging and incorrect action as the main causes of accidents. The question arises of why even experienced professional truck drivers speed and fail to keep their distance, ignoring others’ right of way and priority, i.e. take risks. Spontaneously, one thought that comes to mind is that these types of inappropriate behavior are mainly caused by personality factors. Nevertheless, analyses have showed only slight interactions between the variables of personality and error.


Causes of error

9

Errors

Accident

Fig. 2.2 Three-level model of accident evolvement (Käppler et al., 2008)

The problem of human error came into the focus of several international projects during the nineties, see Reason (1990) and Senders & Moray (1991). The author himself is a member of several international working groups dealing with new methods of in-depth accident and incident analysis based on human error. As one result the Safety Management System (SMS) ARIADNE was developed and realized (Käppler, 2004, 2005, 2006, 2007; Käppler & Dalinger, 2005; Käppler et al., 2008). It may help to get more insight into the aims of driver education and training. The underlying three-level model is shown in Fig. 2.2. Several causal factors (on the left) result in three errors (in the middle) that create an accident (on the right). In Fig. 2.3, these errors are grouped into the three driver activity related categories shown in Fig. 2.1: • Errors of perception are a lack of ability to generate an adequate, consistent image of social and physical environmental characteristics. • Errors of cognition are the formulation of faulty plans based on adequate perception of relevant information. These plans do not follow the requirements of the task or the situation. • Errors of action are the incorrect execution of a correct action plan, mixing up gears or automatically applying similar, common routines. These errors, in turn, are a consequence of causal factors already present in any working environment and technology as latent risks. In Fig. 2.4, causal factors are grouped into eight subcategories.

Errors of perception

Errors of cognition

Fig. 2.3 Error categories (Käppler et al., 2008)

Errors of action

10


Working organization

Communication

Personnel & qualification

Quality Management

Attitude

Physiology

Behavior

Environment

Fig. 2.4 Causal factor categories (Käppler et al., 2008)

• Working Organization causal factors are defects in the purposeful order, regulation and integration of tasks and activities in social systems, e.g. laws, rules or working means. • Communication causal factors are defects in verbal and nonverbal processes of information exchange. • Personnel & Qualification causal factors are the inadequate selection, allocation and qualification of the knowledge and abilities which enable the activity to be carried out. • Quality Management causal factors are defects in the running examination of the quality and state of the work results, e.g. with norms, instructions, manuals. • Attitude causal factors are interferences by workers due to an inclination to value subjects and objects in a certain consistent way, e.g. as a result of craving for recognition. • Physiology causal factors are interferences by workers due to life processes such as growth or illness. • Behavior causal factors are observable interferences caused by workers’ actions and decision-making. This includes processes of experience which are conscious to varying degrees. • Environment causal factors are environmental characteristics which increase risk and can be described on the basis of physical data. Over the past years, more than 2,000 accidents and incidents in road, sea, train and air traffic and the handling of explosives have been analyzed with ARIADNE. The comparison of accident rates has showed the frequencies and types of human errors to be quite consistent in different technical environments. Only minor changes were required to transfer the SMS application from motorbike accidents to explosives handling (Käppler et al., 2008). Consequently, experts today assume that a large number of accidents and incidents are caused by inadequate human behavior and human error. The official accident statistics show that human error is the cause of more than two thirds of road accidents with personal injuries in Germany, see Table 2.1. And it is no surprise that


11

Table 2.1 Driver-related causes of accidents with personal injuries in 1996 (Statistisches Bundesamt; Federal Statistical Office, 2007) Causes of accident

Absolute frequency

Relative frequency (%)

Vehicle operators involved, in total Erroneous behavior by vehicle operators, in total Caused by alcohol Incorrect road use Speeding Distance errors Overtaking errors Right of way, priority errors Erroneous turning Erroneous behavior towards pedestrians Other error caused by vehicle operator

596,982 403,886 19,405 29,495 64,742 47,104 16,120 59,700 33,150 17,791 116,379

100 67,7 3,3 4,9 10,8 7,9 2,7 10,0 5,6 3,8 19,5

almost 20 percent of these driver errors cannot even be explained in detail. This is why the reliability, performance and operation of people, organizations and technology have become a major goal of traffic safety. Whenever human beings act, supervise, steer, control, or make decisions, they affect reliability and safety of the entire system, and any training approach should be based on knowledge about problems and errors of drivers during this process. One additional remark. Table 2.1 hardly deals with behavior, psychological, social or cognitive processes and human errors and is of minor help in the discovery of accident and error causes, because the classification system it is based upon is aimed at settling questions of liability in a legal sense. During this work, the question arose of whether human reliability can be corrected systematically by means of training measures, and which errors can be treated with which measures. Thanks to advanced technology, today’s High-tech simulators may come with impressive quality and realism. They use current driver’s cabs, motion systems which copy the real movements of vehicles to a great extent, and visual systems which generate out-of-window views according to the actions of the driver. The control elements react realistically, noise simulation is deceptively real and the behavior of simulated road users in the virtual driving world also seems closer to reality than one might expect. Simulators like these mean that there are now training opportunities which were previously inconceivable: it is hardly surprising that training concepts aimed at improving safety and human reliability are currently en vogue, and plan to shift about 50 percent of practical training for the truck driver’s license from real roads to simulators, or supplement safety training programs with driving simulators, or streetcar and road driver courses with rail and driving simulators. Many driving schools and other commercial suppliers or training centers offer special simulator-based training programs for driver education and professional drivers, to practice economic and safe driving manners and increase reliability. But which are the basic issues and models behind driving simulation?

12


2.2 Developing a Model Depending on how much practice he has had, a driver’s actions depend on operative and normative concepts which are represented as flexible internal models by the information-processing human being. In the case of practiced sensorimotor tasks, people work multivariately, synchronizing their movements with these models. As they gain experience, the models become automatic to differing degrees. For systematic training, there are various measures which can be carried out, with different steps, to act and make decisions more effectively, faster and safer: See the following diagram according to Godthelp (1992). The training cube shown in Fig. 2.5, for example, shows the three dimensions: • Task: navigation, handling, stabilization • Activity: perception, decision, action • Automation: knowledge-based, rule-based, skills-based. The training modules, of which there are a total of 27, include driving around curves or hill starts. The training concept it is based on assumes that behavior gradually becomes more reliable as it moves from knowledge-based, to rule-based, to skills-based. Hacker (1986) namely believed that targeted action requires temporally stable, invariant representations of: • Aims • The program of action • And the conditions of implementation.

TASK

navigation

AUTOMATION handling knowl edgebased rule-based

stabilization

skill-based perception

decision

action

ACTIVITY

Fig. 2.5 Tasks, activities and automation when driving a car (Godthelp, 1992)

2.2 Developing a Model

13

Accordingly, a targeted action can be regulated in the form of: • Reacting to events • Anticipatory expectation • Bringing events about. Anticipatory expectation means practicing activities which are organized in advance, so that the driver can carry out fast movements precisely. Operative representation systems (ORs) of these invariable process segments are constructed for regulation. The shift is made to anticipation. These operative representations determine reliability and performance. They must develop, at the latest, in the phase of preparation for action, and must remain intact at least until the action is completed, so that the current status can constantly be compared with the target status in a feedback process. ORs are relatively permanent internal models; they are memory representations which guide actions, with both predictive and explanatory functions. Their operative qualities distinguish them from normative representations. For example, they depend upon the type and speed of information processing and motivation, and a system’s navigability or error tolerance. Here, education and training play a major role. If the ORs are not complex enough, activities are based on time-consuming, error-prone, mentally demanding comparisons with external specifications. This is one cause of unreliable actions, delayed interventions, searching for or trying out actions, or of errors altogether. ORs depend upon the demands placed on a person, and represent facts selectively, sometimes distorting them, the scale of this representation depending on their concrete importance. They make generalizations, concentrate on prototypes and only represent classes of characteristics. They are effort-related, leading to strategies for fulfilling requirements with the lowest possible effort. They are encoded in such a way that minimal effort is required to recode between practical execution and the stored model, and correspond with the requirements of the response organization, not only with the information input or the code which is convenient to maintain. For this reason, despite differing information inputs with similar responses, increasingly similar representations develop. The aim is to lessen the load by minimizing cognitive transformation operations (Hacker, 1986). ORs depict targets as the result of anticipations judged in terms of their results; they depict the conditions for implementation which result, for example, from knowledge of how vehicles function, of environmental problems or the probability of accidents and defects. The program of action depicts transformations from the current status to the target status as hypothetical measures to take, with their possible results, and the relationships between the current and the target status, such as the moment to take action and what action can be taken. Depending on how well an OR is learned, a considerable proportion of information retrieval and processing is no longer required (Poulton, 1971). The ideal

14


result of the training course is to shift from perception- to memory-based regulation. This is characterized by time-saving, reliability and a reduction in errors and stress. Orientation is improved when preparing to act, as ORs allow hypotheses to be made about a state of affairs, enabling more efficient search strategies and the selection of information sources which better fit the purpose. ORs have the following characteristics: • They depict facts to match the degree of their importance. They are selective and distorted. • They are of a generalizing, schematic nature and represent classes of characteristics. • They depend upon demand and lead to strategies which enable requirements to be fulfilled with little effort. • They are encoded in such a way that minimal effort is required to recode between practical execution and the stored model. • They form hypotheses, predictions and expectations. Hacker sees the development of operative representations as a multi-stage process which includes a redefinition of the tasks and targets. For example, tasks which require precise numerical values to be derived according to rules are transformed into allocation tasks. Often, representations are encoded more than once. There can simultaneously be both concrete semantic and sensory representations, and different encoding procedures. In the case of everyday tasks, there are signs that more general representations are preferred, and simpler procedures preferred to more detailed ones (Rasmussen & Jensen, 1974). According to Leontiev (1973), to the subject, mental models must appear as a reality which he can be guided by and interact with. Improving the ORs offers opportunities for improving safety, performance and reliability. For example, the task of driving does not only require steps to be taken, but also previous perception and assessment of the situation, planning, checking and repeats as necessary. The modern view of these multiple tasks, activities and conditions aims at improving performance and reliability and reducing errors or the consequences of error. The activities of perception and analysis, planning and deciding, followed by actions, the errors made during these actions and the stress which occurs, are considered the main topics of interest in improving safety and reliability by means of modern teaching and training methods.

2.3 Training Course Design According to Hacker (1986), learning processes determine the representations which develop, not only in terms of their content, but also in terms of their form. One thing which affects their coding is the way information is presented – the teaching matter and methods must be harmonized with one another. Teaching concepts must be designed to encourage the formation of adequate ORs. This requires a knowledge of the ideal ORs to be taught.

2.3 Training Course Design

15

Even carrying out a plan we have made ourselves, behaving reliably according to a specification, is no easy task, as we know from our own mistakes in everyday life. The task of encouraging other people to behave safely is considerably more complex and complicated still. Of course, steps taken to change the environment and traffic neutralize known dangerous situations and make it easier for people to behave appropriately in difficult circumstances. These steps start off with ways of adapting technology and surroundings, i.e. with the vehicles and the traffic environment. Our approach focuses on the active subjects and is centered around errors and causes of error (Käppler et al., 2008). The aim is to help prevent dangerous situations arising in the first place due to individual actions, and to encourage and consolidate appropriate behavior in dangerous situations. Teaching means the acquisition of basic skills, abilities and knowledge by means of targeted measures, with the aim of fulfilling minimum aptitude requirements.

Training builds upon successful teaching and education and aims to sustain and extend existing skills. The mere fact of raising people’s consciousness of issues and topics related to safe driving is doubtless a first step in supporting safe behavior. However, more effective approaches require more than this kind of general verbal conveying of information and making appeals. They use people’s interest in gaining new experiences and repeatedly put them in learning situations. Thus, people are given the opportunity to make mistakes. They experience the consequences of their decisions and actions, and gain experience interactively. They learn that it depends upon them whether or not they get into dangerous driving situations. Of course, anthropocentric teaching and training systems of this kind do not usually constrict people to fixed sequences of actions, e.g. technological constraints. They define teaching aims and criteria to judge success. They offer alternative ways of carrying out tasks by means of variable task segments with room for decisionmaking and action. They are rooted in task- and performance-based analyses of technological and organizational subsystems. Long-term learning results require people to experience real traffic. Theoretical and practical training courses are combined in such a way that there is no longer the conventional divide between cognitive, affective or evaluative, psychomotor components: in real traffic these components interact, too. Using these model predictions and metatargets, teaching and training systems formulate hypotheses about subtargets which are set in an overall concept. Different media, vehicles, simulators and audiovisual systems are used to teach different subject matter. Every teaching device is allotted a clear function and an ideal position in the overall framework of the course.

16


For this reason, simulator instruction alone cannot, of course, guarantee that all teaching targets will be met, but needs to be just one part of a training system, which it is integrated into as advantageously as possible. Training systems of this kind can only be put together with precise knowledge of which types of error occur, where, how and why. A specific training measure thus needs a differentiated analysis of existing weaknesses which is as individual as possible. Figure 2.6 shows a schematized depiction of the design of the training program and its development (Patrick, 1992). It is based on knowledge of the trainee, the training targets and the available training methodology. Of course, courses in driving, or for drivers, cannot actually correct organizational errors or those made by superiors or managers. The subject of interest here is the causes of errors of action and cognition. The author’s investigations into human error show that 60 percent are cognitive, 30 percent errors of action and only 10 percent errors of perception, across different kinds of technological backgrounds (Käppler et al., 2008). As modern safety cultures assume that errors are purely natural products of human behavior, training targets are also not centered around directly avoiding errors by setting up bans, etc. Instead, modern training courses concentrate on the causes of these errors and work on correcting them. For example, Käppler et al. (2008) showed that approx. 45 percent of errors are caused by mistaken behavior, 20 percent by lacking personnel and qualifications and 15 percent by flawed quality management. The remaining 20 percent are distributed among flawed attitude, work organization, communication, physiology or environmental conditions and are still of little

TRAINING TARGETS

Subject matter Tasks Criteria for success

TRAINEE

TRAINING METHODOLOGY

Errors Causes Performance

Means Procedures Teaching strategy

Fig. 2.6 Schematic depiction of training course design (according to Patrick, 1992)


17

interest as corrective targets for costly driving qualification courses, as are the fields of quality management and personnel and qualifications. On the other hand, an analysis of incorrect behavior shows that it is mainly due to deficient knowledge. One example: Motorcyclists have the highest mortality rate, in the field of transport and elsewhere. However, almost three quarters of these accidents are not caused by their own errors, but by other road users. Motorcyclists are not aware that they are often overlooked and that their speed is misjudged (Käppler, 2002, 2005). This analysis points to other correctable aspects, possible training targets and training methods. Long-term traits displayed by drivers include aspects of their personality or education which cannot be influenced by short training courses. In this case, the driver’s current state is of interest, e.g. his motivation or previous knowledge. Knowledge of the driving task or possible accident situations are therefore not a sufficient basis for successful training. Before the training course takes place, information is required not only of qualifications and errors, but also of the trainee’s emotional state, skills, self-assessment, previous experience or social situation. Questionnaires and surveys need to be carried out. Biehl & Brown (1986), Brown et al. (1987) and Duncan et al. (1991) believe that the value of conventional driver training is overestimated, partly due to the lack of this kind of information. This is where the training concept described comes in. The practical examples of application in Chaps. 3 and 4 contain information with respect to this.

2.3.1 Aims and Subject Matter of Training Courses The following training task categories (Käppler & Voß, 1988) show teaching and training topics in conventional literature: • Controlled weather effects, driving in fog, at night • Stress situations, time pressure, emergency procedures • Driving tactics, self-control, evasion, overtaking, driving behind others, advanced driving • Judging distance to vehicle in front, other distances, speed and time • Orientation, navigation, searching, observing, eye movements, eye fixation. These are general topics with varying degrees of practicality which hardly justify the use of expensive training media. This is not the case, however, with operative representations which, because they are used only rarely, are not sufficiently developed, or are forgotten again after they have been learned. To make drivers more reliable, the point will therefore be to correct areas which are stressful, complicated, sensitive, forgotten or unknown. This can be done by speeding up the time scale for the development of ORs and driving experience, using model ifthen-else rules, and for the storage of ways of acting. At the heart of this are

18


Table 2.2 Requirements, contexts and training targets Requirement

Context

Training target

Rare

Never occurs in everyday driving, experienced rarely or only in part Errors of action have considerable consequences Complex technological context

Refresh, cement and ingrain understanding Motivate to strictly adhere to rules Encourage to pay close attention Compensate for problems with understanding by building up familiarity and experience Ingrain simple coping strategies

Sensitive Complicated

Stressful

Fundamental understanding hard to achieve Increased psychological pressure Flood of information Time pressure

Ingrain simple coping strategies

cognitive processes such as awareness, perception, memory, recognition, imagination, thought, conjecture, expectation, planning, problem-solving and decisionmaking (Dorsch, 1987/1991). Table 2.2 shows this kind of rare, sensitive, complicated and stressful situation as requirements for the training program, explains the context in which they occur and lists the training targets.

2.3.2 Training Concept Conventional training courses work from the assumption that people will act safely if they have firm knowledge of the function, context and logic of the system in question. Based on comprehensive knowledge of certain requirements, specific combinations of situations considered crucial for safety are emphasized and specifically drilled. However, there are not enough ways of finding situations of this kind, and very little is known about them. Using simple methods, it is possible to find and describe common combinations which are crucial and occur often in everyday life, but this is precisely why drivers are already familiar with them. Special training in how to deal with these combinations thus seem fairly pointless or even counterproductive, particularly for experienced trainees. Furthermore, experienced drivers generally already have the knowledge they need. It is also unclear: • How long the behavior taught lasts • Which intervals there should be between repeating the courses to effectively stop them being forgotten. Moreover, typical traffic situations involve such a large number of combined elements crucial to safety that specific training seems impossible in the time available. This brings up the question of how to add to conventional training concepts. One argument in particular speaks in favor of this: Most cases of damage take place in


19

Table 2.3 Schematic diagram of the training program Step 1

Situation

Contents

Training target

Stressful Complicated

Psychological pressure Flood of information

Drill simple coping strategies Compensate for problems with understanding by building up familiarity and experience

Sensitive

Time pressure Complex technological context

2

Anticipation

3

Rare

4

Inappropriate

Fundamental understanding difficult Errors of action have considerable consequences Extent to which driver’s behavior is anticipatory Never or rarely experienced in everyday life, or only partly Driving style inappropriate to topography, traffic, resources

Motivate drivers to strictly adhere to rules Encourage them to pay close attention Self-control Analyzing prospective events and adapting actions Refresh, cement and ingrain understanding Experience, reflect on conditions behind situation; learning measures to take

relatively low-stress, generally uncomplicated and perfectly common situations. It is this fact, after all, which always makes people wonder how the accident could ever have happened. Everything points to the fact that the concepts do not cover the entire range of potential conditions for errors to occur, and that there are not enough approaches to stop people making errors due to psychological factors; errors which happen as a result of the situation as a whole. The training program concept presented here is shown in Table 2.3. The first step consists of an artificial, virtual reenactment of crucial safety situations. During this process, a combination is utilized which has been tried and tested in aviation and shipping: creating fixation by repeatedly placing obstacles in the driver’s way. Trainees are tricked into making errors, which they do not realize at first, and they then only experience the consequences, but not how it came about. Following this, the conditions in which the accident occurred and the combination of circumstances are reflected on and discussed, and ways of resisting anger and frustration are demonstrated and practiced. In step two, the idea is to make tangible the extent to which the driver’s behavior is anticipatory, turning it into a focus for closer examination. Here, the starting-point is the consideration that safe actions and behavior necessarily have to consider as wide-ranging an analysis of the situation as possible, looking far into the future. In step three, situations are drilled which are rarely experienced, or which are known only by word of mouth, and which involve large potential risks, e.g. tractor trailer brake failure when driving downhill, see above. In step four, situations are practiced which require the driving style to be specially adapted to the topography or the traffic, e.g. in order to save fuel.

20


Finally, one point must be brought up which is often neglected: evaluating the training course itself. This requires a comparative analysis of accident and damage frequency in trained and untrained groups of drivers. In the form of a kind of transfer test, the effectiveness of all scenarios in the simulator must be tested, and individual scenarios fine-tuned. A scenario in which nearly all drivers manage to deal with brake failure, for example, without any problem, can no more be considered a success than can obstacles set in their path to no effect. The fine-tuning, including that of the situational circumstances, requires means of testing which can capture and analyze data of this kind. The type of events and their integration in a context of higher-order tasks and targets are of great importance. Importance is attached to ruling out the isolated observation of “good” features, e.g. oriented toward technological feasibility, or arbitrary combinations of tasks and events: see the practical applications in the following chapters. Simulation requires the conception of a training system which covers the vehicle, simulator and other means and systems and is based on a detailed, systematic analysis of aims and requirements (Roscoe, Jensen & Gawron, 1980).

2.3.3 Description and Analysis of Activity So what is actually drilled? At the heart of practical training courses are tasks which pursue the above-mentioned targets and are typical of processes in real life. They are found by analyzing tasks and activities, aims and boundary conditions, as well as technological and organizational elements. This results in creating tasks and setting down training methods and media. These are based on classification schemata or taxonomies of certain aspects of an activity. There is a long tradition of developing this kind of taxonomy for a large range of different activities and applications. The main aspect of selecting a certain taxonomy is the practical application of its results. The steps toward developing training units are: • Describing the task and activity • Analyzing the task and activity • Identifying problem zones and critical incidents. Often, descriptions of the task already exist and can be analyzed. Here, it must be taken into consideration that the actual implementation of the task is highly likely to differ from the description of how it should be carried out, and indeed generally does. Thus, in the first working step, it is important for the activity to be described. This has the advantage of describing the activity which was actually carried out. The descriptions of the task or activity are then analyzed, using information (already implicitly contained in the description) about the contents, processes, means of assessment, measurements, boundary conditions etc.


21

An understandably brief overview of task analysis processes can be found in Patrick (1992). Below is a short depiction of a well-established process by Kirwan & Ainsworth (1993): task decomposition. As an example of application, an approach by Goettl (1993) is described for systemizing the development of sensorimotor skills in a flight simulator. Basic skills are identified using the reverse transfer technique. Basically, this means decomposing complex flight tasks into simple subtasks (task decomposition, Gopher et al., 1989). The theory behind this sees expertise, like Hacker’s ORs, as an organized pattern of reaction strategies which are implemented flexibly. Task decomposition combines elements of skills theory with attention theory and is based on similar structures and elements between simple subtasks which follow one another. Let us continue with Goettl’s example, where a flight task is learned by starting off simply flying straight ahead at increasing speed. Building upon this, the loop learning situation follows, first in two and then three special dimensions, also at increasing speed and with decreasing radii. The complex situation which follow consist of a low-altitude flight through a landscape of buildings with a high degree of vertical structuring. The special feature of Goettl’s procedure (1993) is that similar elements are identified, using performance criteria, by means of correlations between the complex low-altitude flight and the individual prototype situations. The aim is to speed up learning by transfer from simple situations at the start to subsequent situations which increase in difficulty. Extra time is taken when it takes longer to learn complex task structures than to combine simple tasks which were previously separated. Papers have been published on task decomposition since the end of the ’80s. There are no known applications in driving simulation, but they can be derived in analogy to aviation. For more in-depth information, see papers on work analysis processes and situation definition (e.g. Küting, 1986; Fastenmeier, 1988a; Jensch et al., 1978; Schagen et al., 1987; Utzelmann, 1985; Frieling & Hoyos, 1978). Descriptions are not available, or even possible, for all tasks, let alone the activities. One example is critical driving situations and accidents. Here, it helps to centre on the skills required to process tasks safely. The tasks themselves are analyzed; the necessary skills and means of assessment are defined. Using this information, specific training courses are composed. This process initially appears simpler than it is to carry out. The main problem is ascertaining which skills are actually necessary to process a task safely – a job for experts. Guilford’s (1977) corresponding structure-of-intellect model shows clearly that even a classification of cognitive skills is already bewilderingly interlinked. During ten years of research Guilford himself extended the number of factors to be taken into account from the original 120 (Guilford, 1967) to 150 (Guilford, 1977), see Fig. 2.7. The statistic rotation system of the main component analysis can be criticized, as the choice of factors is based on subjective decisions. Thus, Guilford’s model can be seen as a rich source of hypotheses on the nature of intellectual activities and can contribute to theoretical understanding.

22


CONTENTS

visual auditory symbolic semantic behavioral

PRODUCTS

units dasses relations systems transformation

OPERATIONS

illusions

evaluation convergent production divergent production memory cognition

Fig. 2.7 Structure-of-intellect model (according to Guilford, 1977)

2.3.4 Evaluation and Effectiveness Duncan et al. (1991) proved that without effective structured feedback, gaining experience does not necessarily lead to the driving experience desired. One central argument for using simulators is that driver-specific interactive data can be acquired and evaluated, to provide feedback on and assess driving and operating quality. The poor cost-effectiveness ratio of the simulator compared with a vehicle makes it more or less obligatory to use this data, which, in reality, cannot be captured as extensively, in such detail, and above all in such a well-regulated and standardized manner. During training, this data can be used to demonstrate very specifically which behavior by drivers is good or needs improving. Driving style can be determined using suitable assessment methods. There are already some performance criteria, or at least some experience has been gained in experiment, related to the evaluation of driving quality, but most criteria still need to be elaborated. This is particularly true of safe driving. However, as well as appropriate parameters, it also depends on the quality of the combinations of situations used, and the task itself. For this reason, appropriate parameters need to be determined not only to evaluate the driver’s behavior, but also to assess the effectiveness of the training situations themselves. The procedure is as follows. While the training program is being tried out, after the first data has been gathered on good and poorer drivers, e.g. in terms of

2.4 Training Media

23

fuel consumption, relevant parameters must be separated from irrelevant ones by a statistical comparison of these extreme groups. These values are used to assess combinations of situations. Those combinations are selected which produce greater increases in learning than the others. The influence of specific personality variables can also be checked, such as age, driving experience or the type of vehicle used every day. During the evaluation process, data from good or poorer drivers are made into performance criteria. These must be reliable, simple to understand and easy to interpret. The prerequisites are knowing about and having a clear vision of learning targets, tasks, values to be achieved and the resulting interventions and training measures. Hence, “artificial values” which are complicated to calculate and almost impossible to interpret are given a wide berth.

2.4 Training Media The subject matter of the course produces technological and operational specifications for teaching and training media. All-round orientation, looking to the rear or increasing fixation distance to improve safe following distances can all be practiced using appropriate audiovisual methods, independently of the driving process. Basic skills training in part-task trainers integrated into full-mission situations was already suggested by Gubser & Spörli (1970), as it clearly has advantages in terms of acceptance. By contrast, demonstrating the influence of alcohol on driving ability, while popular, is problematic, as demonstration without real consequences can result in false conclusions. Table 2.4 shows the media available within the context of driver instruction and training. Table 2.4 Teaching and training media, tasks and examples Medium

Type of task

Example of task

Vehicle Full-task simulator

Practical driving exercises highly dynamic, emergency events

Normal driving simulator Procedural simulator Part-task trainer

Quasi-stationary normal trip

Realistic Brake failure Evasive maneuver Motorway, secondary road

Computer-assisted instruction/training, (CAI, CAT) Tutorial systems Audiovisual Paper and pencil

Learning tasks, rules, procedures for action

Navigation through a city Hill start Stopping Introduction to simulator or task

Introduction Lessons Test

Introduction to software Theoretical lessons Examination

Simplified procedures in trip as a whole Isolated tasks

24


2.4.1 Simulators This manual is about the use of simulators as modern training media to improve reliability. Why simulators? As early as 1929, Edwin A. Link introduced his Pilot Maker, a blue airplane cabin on a moving platform. He developed the Total Training System, which, apart from flying instruction in the simulator, involved only two hours of flight in an airplane, and guaranteed people would learn to fly for US$ 85. In the first year alone, 100 pupils left the Link Flying School as trained pilots. In 1934, Link managed to sell six trainers to the US Army Air Corps – and the simulator industry had been founded. In the Second World War, more than 500,000 Allied soldiers were trained in more than 10,000 blue boxes. Modern pilot training can hardly be envisaged without flight simulators. Lufthansa, for example, carries out more than 90 percent of commercial pilots’ practical training on the ground. At the end of the 1950s, industrial societies had visions of infinitely plannable, feasible and controllable technical and social progress; they had optimistic expectations of what was technologically possible (Geißler, 1992). Even experts believed that road traffic safety could be improved by means of driving simulators. Thirty years after Link, in 1960, the University of California in Los Angeles introduced the first interactive driving simulator. It was for scientific research and consisted of a movable seat in a driver’s cab, a projected picture and simulated force feedback (Hulbert & Wojcik, 1960). Over the 1960s, ’70s and ’80s, many different driving simulators were developed for totally different scientific and training purposes. At the start of the 1990s there were already 50 verifiable driving simulators of differing construction around the world (Käppler & Alexander, 1995). Altogether, however, the improvements and success achieved by driving simulators remained well behind expectations. According to Evans (1991), despite considerable investment – Daimler Benz actually put 26 million DM into one single driving simulator at the start of the ’80s – there is no discernible contribution of scientific interest causally linked to the use of driving simulators. Simulation is the simplified replication of any system or process by means of another system or another process, and experimenting with this model. This view is still widely shared today, with only few exceptions. Not only the scientific community, but also industry and investors, are asking for the reasons why – and there are many reasons. Simulator-assisted training is a thoroughly interdisciplinary specialist field. When combining such different fields as informatics, psychology, educational science, or electrical and mechanical engineering, difficulties occur, and not only due to the different nature of these fields, the ways in which they overlap or attempts to keep them separate. Depending on how you see it, simulation is a science, a technology or an application. Semantic problems also play a role, and it is possible that communication problems contributed quite considerably to the difficulties. For this reason, first let us answer the question: What is simulation?

2.4 Training Media

25

2.4.2 Definitions The Latin simulare means to imitate or to copy, but also to pretend, to play a part and to feign (Petsching, 1967; Meyers, 1987). Simulation is therefore playing a part, pretense, sham; the Latin plural means the art of deception. (Dorsch, 1987/1991) adds to the definition, calling it a simplified representation of a slice of reality, and points out that the models used can be of a physical, technological or abstract nature. This is the case with numerical simulation. It is the recreation of systems by means of mathematical equations on digital computers. The computer programs it is based on take the form of theoretical models and allow combinations of variables and possible solutions to be played out. The aim is to understand these processes and the factors which influence them better, for example in testing scientific theories or in planning mechanical or administrative systems. In reproducing tasks whose accomplishment by human beings would be described as an act of intelligence, computer simulation has broad overlaps and similarities with artificial intelligence, see Lotens et al. (2005), e.g. When the time taken by the processes depicted remains unaltered when reproduced on the computer system, this is called a real-time system. In this case, the computer simulation is subject to quantifiable schedule requirements designed to synchronize events. Real-time systems can be found, for example, in automation and control engineering, or in informatics. The use of the term “real-time system” occasionally appears generous. For example, the financial services provider Vereinigte Wirtschaftsdienste (United Economic Services) calls its information system on current stock exchange prices a real-time system because the prices are transmitted constantly during the trading session. Hewlett Packard calls its PostScript process for calculating edge correction in laser printers a real-time algorithm as it blocks printing until these calculations have been carried out. Account management systems in banks and airline dispatch systems are also described as real-time systems. Here, the customer himself chooses when to use his account and synchronizes the process, and only has to wait a short time. The timing is flexible. In contrast, the timing of true real-time systems is inflexible, involving precisely fixed times. In production processes, for example, individual goods are conveyed on belts and taken to further processing steps by robots. By the time a workpiece arrives on the conveyor belt, the robot must be ready to pick it up. This synchronization is ensured by means of runtime control or buffering. Control mechanisms in nuclear power plants work in a similar way. They measure the reactor temperature with a repeat rate of 10 Hz and, if the maximum temperature is exceeded, set off the shutdown mechanism within 50 ms. If the shutdown did not occur exactly on time, and the cooling system experienced total failure, emergency shutdown could no longer be guaranteed. With these true real-time systems, falling behind schedule produces expenses which are not incurred when the schedule is followed. As a result, staying on time is an operational target which can be achieved by keeping to deadlines rather than by other synchronization measures. For computers, these deadlines are the start

26


and finish times: the earliest point a computation can start or the time it must finish by. Schedule requirements are generally focused on keeping to target finish times. To separate driving simulations from these true real-time systems, the term real-time simulation is used in the literature. Real-time simulations are used for demonstration, experiment or training purposes, such as pilot training (Dorsch, 1987/1991; Meyers, 1987). Sonntag (1989) describes them as a special kind of cognitive training and investigation process. Real-time simulation does not apply to the simplified depiction of processes which can be objectively determined by others. It is used if, when practicing tasks in real life: • • • •

Falling behind schedule or making mistakes leads to system failure Practice is seldom possible The task cannot be reproduced due to the dangerousness of events in real life The task cannot be reproduced due to the structure and complexity of situations. Real-time simulation means a human/machine system which uses one or more media to depict information. This information can be in the form of optical, acoustic, haptic or vestibular processes, which can be represented by means of film, a computer screen or a servo motor and springs. One characteristic of real-time simulation is direct interaction by one or several people with this information for the purpose of information processing. Its aim is to create a simple reproduction of a system or process, or slice of reality, using another system or process, and to use this to stimulate subjective representations of the environment, so that people have the illusion of fulfilling real tasks.

In this way, the range of applications for simulations corresponds to the training targets for increasing reliability, described above. This is why real-time simulation is so widespread in airplane operation. To train pilots, the environment is recreated, so that pilots are subjected to less danger, even in critical situations (Frieling & Sonntag, 1987; Drosdol & Panik, 1985; Flexman & Stark, 1987). Now it becomes clear what a simulator really is, in the original meaning of the word: a person, an imitator, or, in the words of Socrates, a master of the art of deception (Petsching, 1967). In our application, it is about replicating movement from place to place: the task of transportation is not carried out at all. The simulator imitates conditions so realistically that drivers experience the unbroken illusion of dealing with tasks. The main features of simulators, therefore, come from requirements determined by the ways people take in and process information. Investigations by Springer et al. (1993), for example, showed the significance of visual, tactile and proprioceptive information. Constantly providing this information, on time, continuously and without delays, is one important operating target for simulators, as unless the flow of information is constant, the perception of moving through space is disturbed.

2.4 Training Media

27

One example: A driver turns the steering wheel when entering a curve. If the representation of the out-of-window view, showing the motion of the computersimulated vehicle, turns with a delay, e.g. because the computer calculates too slowly, then the steering degree cannot be adjusted on time. A simulator is a device which enables people to operate vehicles with no real movement, but in realistic conditions. A virtual vehicle of this type can hardly be kept on the virtual road at high speeds. Practicing the task of entering the curve is only possible if special skills are learned for using the simulator which, in the original situation, would be unnecessary or even a hindrance – such as prematurely setting an exaggerated steering angle in anticipation of a curve. Teaching and training are pointless or counterproductive, however, if learning processes from the first, simulated test situation cannot be transferred to carrying out the second, real test. This transfer effect can have a positive influence on processing a task, or, under certain circumstances, even have a negative influence. This is known in the literature as positive and negative transfer (Latin transferre, “bear across,” Dorsch, 1987/1991). As a result of current advances in computer technology, more and more new techniques are being created for virtual representations of the environment, e.g. the use of input and output media such as data gloves, or graphic representations using helmet-mounted displays or virtual-reality goggles. In view of this, the question arises of which requirements real-time simulation has to satisfy in order to comply with teaching or training tasks with maximum possible transfer. The answer to this question goes beyond technological requirements for simulators as regards simply staying within schedule when delivering information. In fact, the operating aims of simulation are subject to a range of other requirements, whose common aim is to enable the best possible transfer from the test situation to real life.

2.4.3 Developing a Model, Transfer and Validity Transfer answers the question of the extent to which dealing with one task makes processing further tasks easier. If totally new activities are required, transfer does not usually occur. Müller (1911) discovered early on that the transfer of learned skills from the learning situation to application does not take place generally, but is restricted to similar learning matter and learning methods. Bartlett (1947) investigated transfer between identical classes of specific learning situation and stimuli. He found that one condition for positive transfer is that the learning situation must be harder than the transfer situation. However, Nissen & Bullemer (1987) and Braun et al. (1993) showed that complex processes are learned faster when the stress is reduced. A differentiation must therefore be made between learning conditions.

28


According to Thorndike & Gates (1930) transfer occurs between situations containing identical elements. Thus, practicing adding up should have a positive influence on learning to multiply, in that multiplication is seen as combined adding, and the latter recognized as an identical element of the former. It has, however, been pointed out that the search for identical elements could be never-ending, as identical subcomponents must be found which are increasingly harder to grasp. Thorndike is thus not able to determine the exact nature of identicalness and to explain to what extent conclusions can be drawn or generalizations made about the task as a whole, based on the learned element (Pavlov, 2006). In view of the difficulties and doubts involved in hierarchical decomposition, even of simple tasks, into clearly defined subcomponents, the assumption about identical subcomponents seems dubious. Bergius (1964) therefore dealt with structural conditions for transfer and was of the belief that transfer is most likely when insights are gained into circumstances and procedures. At the end of the 1980s, transfer was still seen as the central problem of all training applications, even of real-time simulation, e.g. by Hale & Singer (1989). They believed simulation had to deliver benefits which were known and plannable in advance (Blaauw, 1982; Häcker, 1972; Roscoe & Williges, 1980; Käppler, 1997a). Otherwise, they believed, experimental investigations, et cetera, would be pointless, or teaching and training even dangerous. Transfer was considered the aim of simulation. Transfer occurs when a task set is new, but the way to deal with it requires similar activities to the original situation. Transfer is strongest between very similar situations and becomes weaker as the similarity lessens.

Validity deals with statements on the soundness of tests, in the sense of how precisely one criterion can be measured (Dorsch, 1987/1991) and shows the degree of precision with which a test actually measures that trait or behavior it is intended or claims to measure. This is demonstrated by Fig. 2.8, below, which shows that transfer in simulation is not an insignificant process. From everyday life, we know that there can be little correlation even between what is known as the objective event (circle on left) and our subjective view of it (circle on right). Pirandello (1999) came up with a good phrase for this: “It is so, if it seems so to you.” If we construct this, our subjective view, mathematically, as a formal incident or event model, e.g. of how the accident occurred (circle at top), in our example the model validity only covers conditions in the upper triangle of the ellipse, somewhat shaped like a Rotary engine rotor. A mechanical simulation of the event (circle at bottom) is then added to our example and the mathematical model is extended, making it a driving simulation. The conditions of the investigation are varied dynamically and the drivers are trained and studied. This example shows clearly that the validity of the simulation results

2.4 Training Media

29

Fig. 2.8 Model development and validation Validity

Incident Model

Objective event

Model validity

Subjective view

Simulation of event

Correlation

is guaranteed when none of the conditions, views or simulations reach beyond the boundary of the “lens” in the centre of the picture. Minsky (1988) shed light on this in a good, ironic example, see Fig. 2.9. He explained as follows: Model M1 is a model of object A for observer B1 to the extent that he can use M1 to answer questions which interest him about A. On the other hand, for observer B2 , model M2 is a model of object A to the extent that he can use M2 to answer questions which interest him about object A. With this knowledge, Binninger (2000) formulated his succinct, and also slightly ironic, “Laws of Simulation,” which speak for themselves: • • • • •

DON’T believe your model is reality DON’T extrapolate beyond the region of fit DON’T distort reality to fit the model DON’T retain a discredited model DON’T fall in love with your model.

M1

M2

B1

B2 A

Fig. 2.9 Model development according to Minsky (1988)

30


Thus, the logical procedure with simulation can be summed up as follows: • The aim of the simulation needs to be set down in a technological, scientific hypothesis. • Models must be developed and evaluated, or selected, as an abstraction of reality. • Simulators must be developed or selected as a means of carrying out training courses or lessons. • Boundary conditions must be set for the simulation. • Boundaries must be set for the validity of the results. • The validity of the results must be proved. What is therefore needed? • A well thought-out concept • Representative subject matter for the investigation and the training course • An experimental environment for training. This set of tasks and difficulties is accompanied by linguistic ambiguities. Patrick (1992) complained, for example that fidelity should be the theme of every discussion on simulators. The term fidelity comes from information theory and means the property of diagnostic tests to provide very precise measurements which are not, however, very wide-ranging. The term can hardly be applied to the transfer of learning processes. Patrick (1992) does not go into these questions at all. Steininger (1995) admired the precision and faithful reproduction of almost perfect training devices: Using digital simulation, he said, it has been possible to improve the fidelity of the simulators so greatly that almost total simulation has become possible, rendering earlier questions of transfer of training superfluous. The author is not able to share this optimistic attitude. Even if simulators are perfect, this still says nothing about whether or not their use makes sense, about the correspondence between training targets and the overall context, the representativeness of training tasks or the validity of evaluations, etc. It is precisely the high technological perfection of the training media which tends to obscure discussion of these questions, and the corresponding answers (Mehl & Käppler, 1998a). Thus, simulation, in particular, needs to ask itself the question of how valid the training results are for the intended purpose. According to the Standards for Educational and Psychological Testing (AERA, APA, NCME, 1985), three types of validity can be distinguished (Anastasi, 1990): • Construct validity • Content validity • Criterion-related validity. The aim of construct validity is to provide a theoretical explanation for results; it is achieved by using logically structured concepts for teaching and training. It is comprehensive, as it focuses on explaining the causes behind circumstances.

2.4 Training Media

31

The aim of construct validity is to provide a theoretical explanation for test results. It deals with attempts to explain the meaning of a measurement by means of the theoretical background. It is achieved by using logically structured concepts, e.g. for teaching and training. It is comprehensive, as it focuses on the formulation of hypotheses, i.e. explaining the causes of circumstances.

Content validity is related to the test contents. It occurs when the test tasks already contain the trait to be measured. This special case is described as logical validity. The content validity of the simulation implies that the test situation and subject matter are representative of real situations, and that registered measurement values are actually performance criteria, e.g. of learning outcomes.

Criterion-related validity is the quantifiable or empiric validity. It is related to statistical relationships between test values and criterion values, which are placed in relation to one another.

The content validity includes the face validity. This relates to what a situation seems to depict, at face value. In itself, face validity cannot replace objective validity. High face validity does not improve the content validity of the simulation. Its advantage is that it improves the acceptance, motivation and cooperation of the people involved. This leaves criterion-related predictive validity. The time the test value and criterion value are ascertained means that it makes sense to differentiate between concurrent validity and predictive validity. During the training course, the test values should predict future criterion values. Thus, transferability predictions result in a quantification of the predictive validity of results. Strictly speaking, validity requires the test creator to inform the trainee in advance of the required test results. Indeed, determining the predictive validity for every specific learning situation is problematic, if simulators are specifically designed to be used in rare, dangerous and highly complex training units. Precisely for this reason, comparative data, e.g. for critical driving situations, cannot be produced at all – and these slices of reality are precisely what the simulation changes, making them common and harmless, with no consequences if errors are made. However, if this is the main characteristic of the simulation, the requirement for validity becomes extraneous. The content validity of these slices of reality is not possible at all, as they have been changed into central characteristics, and are, for instance, harmless. The question arises of the extent to which results of this type can be validated at all.

32


But if simulator training is integrated logically and sensibly in an overall system, construct validity can be assumed. There is content validity when educational situations are representative of real situations, and when measured values are actually performance criteria for learning outcomes. Face validity guarantees the test subjects’ acceptance, motivation and cooperation. One approach to criterion-related predictive validity is comparing assessments of driving quality in the driving simulator with corresponding assessments in the real-life situation. To do so, a representative sample of the driving task, situation, vehicle and drivers must be investigated, and these estimated values collected and compared with those from the corresponding real situation. In this way, criterionrelated validity can be quantified, although with subjective estimated values. Of course, correlations between performance criteria in the simulator and the vehicle can also be calculated and checked for their significance. This is probably also the reason for Steininger’s (1995) optimism. Hollnagel (1981) even took the view that criterion-related validity is not necessary at all, if there is enough construct and content validity. In this case, he believed, training results could be transferred qualitatively to the real situation.

2.4.4 Driving Simulators: Setup and Requirements

Variability

Figure 2.10 is an attempt at distinguishing between different simulation processes. It shows the variability of a development or investigation tool and the quality of the results it aims at, labeled the degree of realism (note: not the validity). Although its variability is higher, a constructive simulation with a sketch or numerical simulation provides only little reality. The whole system, as well as the operators and tasks, are

Degree of realism

Sketch

Simulator

Fig. 2.10 Structure of simulation processes

Real system

2.4 Training Media

33

all simulated. In the center, virtual simulation includes driving simulators. Here, the operator is real, and the vehicle and task still simulated. In the so called live simulation, the system and operators are real, while the task may still be simulated, as before. Due to actual simulator design and requirements, either with high variability or high realism or a good mix of both, simulators may be put almost anywhere in the middle of that Figure. Which characteristics place a driving simulator where? Figure 2.11 shows the schematic layout of a driving simulator. It consists of a driver’s cab, known as a mock-up, often original vehicle or replicated bodywork or parts of bodywork, and other simulator element, controls and simulation of force feedback: • • • • • • • • • • • • • •

Mock-up Vehicle model Traffic model Road model Visual model Sound model Motion model Force feedback model Display model Data recording model Data evaluation model Replay model Training model Courseware.

The mock-up contains all the relevant controls and displays. The driver’s operation of the controls is registered electrically and transferred to computers. In the

Vehicle model

Display model

Traffic model

Data recording model

Road model

Data evaluation model

Visual model

Replay model

Sound model

Training model

Motion model

Courseware

Force feedback model

Fig. 2.11 Schematic layout of a driving simulator

34


digital computers, programs are installed for simulating the vehicle dynamics. From the gas and brake pedal input parameters, and the position of the gear lever, for example, they calculate the driving speed. The output variables for the vehicle dynamics are used to control programs for the out-of-window view. It creates an out-of-window view from the driver’s perspective on the screen of a graphics system or using a projector, depending on the vehicle’s position and turning angle in the data base. Furthermore, other information is displayed, such as the speed on the speedometer. Other simulator elements replicate engine, tire and wind sounds, as well as the force feedback, such as the steering momentum. Motion or acceleration simulation recreates the vehicle’s movements across the terrain: for example, the mock-up is moved using complicated equipment. The main input parameters for this simulator element are the vehicle movements calculated in the dynamics model. Another simulator element recreates the steering momentum, depending on the steering angle, the lateral acceleration and the speed. A driving simulator must fulfill certain minimum requirements to serve as a useful training medium for the target training purpose. The quality of the simulated vehicle dynamics, with the movement and out-of-window view, have a decisive influence in conveying the effect of a road trip. There are vehicle models which provide a close approximation of driving characteristics in many ways. However, as regards their programming technology, they are extremely elaborate, and running them in real time on a process computer is problematic. An out-of-window view with a lot of detailed information is desirable from the point of view of the driver, but processing the necessary volume of data in real time can only be carried out at great expense. Over the years spent dealing with simulators, expert knowledge has led to a wide range of technological requirements for them. They mainly revolve around creating the illusion that the test subject is carrying out a driving task in the real world. If stimulus-response patterns in stimulators are similar in quality to those in the real world, they can be perceived and integrated in this way, and processed to produce decisions. Representing the information on time and without delay is one of the most important operating targets for simulators, as unless there is a continuous flow of information, the perception of moving through space is disturbed. Overviews of out-of-window view specifications, now outdated, can be found in surveys of the literature by Padmos & Milders (1992), Haug (1990) or Korteling (1991). A group of specialists from the US National Research Council (Käppler et al., 1992) compiled specifications for driving simulators and 54 driving tasks. The German Federal Highway Research Institute (BASt) had a report written up on simulators’ suitability for training and testing drivers (von Bressensdorf et al., 1995), with a summary of the most important technological characteristics; this is still fairly up-to-date. Altogether it can be seen that technological requirements for driving simulators are very demanding, due to the high angle-of-view speeds, high-frequency system dynamics and the driver having a lot of freedom. In general, a specification is only valid in the context of the task it is derived from, meaning that there is no one single specification for the driving simulator. There now follows a short discussion on the main characteristics from von Bressensdorf et al. (1995) and Käppler & Alexander

2.4 Training Media

35

(1995). Interested readers can find out about technological requirements here and note that these are still essentially dictated and limited by technological development opportunities – and development is ongoing: ideas which open up new opportunities one day are dismissed again the next.

2.4.4.1 Out-of-Window View Requirements in terms of the out-of-window view dictate, among other things, that visual information must be provided smoothly, without delay and in real time. Their quality can be seen from the following characteristics: • • • • • •

View to front and back; visibility Horizontal and vertical resolution; luminance Information content Segmenting and levels of detail Frame rate and transport delays Textures and anti-aliasing.

For traffic tasks, 180-degree horizontal angles of view were once considered sufficient (Padmos & Milders, 1992; Käppler et al., 1992); today this is already far too little. The visual range, especially in city traffic surroundings, must extend to below 1 m (3 ft), so that it is possible to drive close up to traffic lights, for example; in many cases, this is not possible. The maximum visual range is limited by the computing capacity and the resolution of the display systems. The human eye has a minimum resolution of 3 arc minutes. Thus, modern monitors’ raster technology require 15 lines per degree; to identify traffic signs, 60 lines per degree are necessary (Padmos & Milders, 1992), but can hardly be achieved. The literature describes luminance of 20 cd/m2 as sufficient, but these values are minimum requirements at best. Segmentation cuts a data base into adjoining sections so that the entire course of the road ahead does not have to be calculated from the terrain in the data base, but only for the section in the visual range in front of or behind the driver. This can lead to pop-up effects during fast driving, when large segments of street pop into view at once, as well as levels of detail. The latter means the degree of detail in displaying objects which are integrated dynamically, depending on how far the observer can see. The impression of moving along is created by presenting individual images in fast succession. From a frame rate of about 30 images per second, human observers start to perceive it as smooth movement. In general, higher frame rates are needed, in the interests of high image quality and faster vehicle reactions, and to avoid visual elements popping up into the peripheral field of vision. This is just as important a requirement as that for smooth-running repeat rates, to avoid misleading motion information from images which are sometimes smooth, sometimes jerky. Transport delays (the delay between the driver’s input, e.g. with the steering wheel, and the display of the corresponding out-of-window view) should preferably be below 60 ms in the interests of smooth systems dynamics, but this can hardly be achieved.

36


2.4.4.2 Motion, Display, Controls and Sound Specifications for the layout of technological systems to replicate acceleration, motion or vibrations concern distances, angles, acceleration, transport delays and the corresponding cut-off frequencies, and none have been validated. Thus – unfortunately – it remains up to the individual simulator designer’s expertise to make a good choice, or determine one by trial and error. There is still reason to believe that here, especially, the cost/benefit ratio is very unfavorable. Replicating force feedback and sounds ensures that the vehicle can be controlled realistically, and that only short training sessions are required to get used to it. Here, once more, the manufacturer and designer are mainly left to rely on trial and error due to lacking or poor specifications. 2.4.4.3 Vehicle Dynamics In simulators, for reasons of capacity, the number of vehicle parameters available for variations is considerably lower than in real vehicles. Tire characteristics are often linearized and models simplified. This does not have any negative consequences as long as the simulations are limited to their expected uses. In the case of the one track model, a linear model with three degrees of freedom, for example, the wheels on one axle are seen as one imaginary single wheel. This model can yaw and move in a longitudinal and lateral direction, but it cannot properly roll, pitch or make vertical movements (oscillations). Thus, linearized models may only be used for small drift angles; the maximum permissible lateral acceleration is then considered to be 0.3 g. Dynamic behavior in the borderline area ought not to be practiced with simulators of this kind . . . Complex four-wheel models, on the other hand, can replicate all degrees of freedom. Here, ultimately, the number of parameters specifies the range of variations and the precision of a replication of vehicle dynamics. For this reason, there are often immense differences in quality, and the parameters – unfortunately – are often not even validated, but adjusted according to taste. The problems then start when training with experienced drivers . . . 2.4.4.4 User Support One very important requirement is that the simulator and courseware are useroriented and simple to use. This relates to simple-to-use ways (even for people other than computer scientists) of selecting and adjusting: • • • • • •

Vehicles, trailers and loads Environmental conditions Traffic conditions Events and what they feature Levels of difficulty Measurement values.

2.4 Training Media

37

The composition of tasks, routes and events must be user-friendly and provide assistants for navigating through the scenarios. Sufficient information must be available during the simulator training session. For the evaluation, it must be possible to select performance criteria and illustrate and process them quickly and simply after a trip. The usage examples in the following chapters show that this is all in development.

2.4.5 Typology of Driving Simulators Another important point has also not been dealt with. When can a driving simulator be called by this name, and what characteristics must it then possess? Can any driving simulator be used for any teaching or training? Of course, one driving simulator is not like the next. There is a whole range of different constructions with different qualities and purposes. This subchapter proposes a conceivable means of categorizing the technology and how it is used; see Fig. 2.12, including examples. The classification proposes six different categories and is based on known needs in terms of human perception when driving, i.e. mainly the depiction of movement and the out-of-window view, above all. It must not be forgotten that this is an attempt at a technological classification of simulator construction, which does not say anything at this point about how they are used on driver training courses. Corresponding

2

3

1

4

6

5

Fig. 2.12 Technological typology of driving simulators in six categories

38


requirements, rules and guidelines need to be developed when firm information is available on teaching and training outcomes. The current state of knowledge is not yet sufficient for this.

2.4.5.1 Constructions with no Degree of Freedom We start off on the overview in Fig. 2.12 at 9 o’clock, with the most simply constructed static fixed-base driving simulators with no degree of freedom. Both mockups, number 1, and reconstructed or modified bodywork, number 2. Below, Fig. 2.13 shows the schematic representation of number 1, a static driving simulator with a mock-up cab and primary controls, with zero degrees of freedom (dF = 0).

1

Fig. 2.13 Schematic representation of a static driving simulator with a mock-up cab and primary controls (dF = 0)

These simple driving simulators are characterized by: • Optical representation of dashboard and surroundings • Acoustic representation of surroundings • Haptic feedback from controls. As they lack any form of movement, and, especially, as they are similar to toys, driving simulators of this type are not well suited to serious training courses. These types are mainly in use for demonstrations or simple part task research or training applications One example which has been put into practice is the police pursuit trainer in Fig. 2.14, the product of cooperation between Time Warner and Atari Games Corporation AGC at the Technical University of Berlin. Figure 2.15 shows the schematic representation of number 2, the same version with a modified vehicle instead of the mock-up cab. The advantage of this is the face validity with realistic vehicle shape for the field of view, and the sounds. Here, again, the disadvantage is the lack of any movement for the driver. As an example in practice, once again, Fig. 2.16 shows the IVI Driving simulator of the Fraunhofer Institute for Transportation and Infrastructure Systems in

2.4 Training Media

39

Fig. 2.14 Time Warner’s AGC SV500LE static simulator at the Technical University of Berlin with mock-up cab, primary controls and out-of-window view on 5 screens (dF = 0)

Fig. 2.15 Schematic representation of a static driving simulator with modified vehicle bodywork and primary controls (dF = 0)

Fig. 2.16 IVI Driving simulator of the Fraunhofer Institute for Transportation and Infrastructure Systems in Dresden

2

40


Dresden; a close to series vehicle without motion and extended outside view. It is used for research into driver assistance systems, e.g.

2.4.5.2 Constructions with up to Three Degrees of Freedom This construction can be extended, for example by adding simple movement with information on the shape of the road surface. The advantage of this is feedback on road cues in a vertical direction, and the haptic information. Such simulators may be equipped with pneumatic pistons, with affiliated coil springs and shock absorbers located in each of the vehicle cab’s wheel wells. This system simulates the normal vibrations experienced while driving and can provide minimal car cab pitch. The top version of these constructions adds other movements apart from road surface cues and has three degrees of freedom at most. One special feature is that the vertical vibration is imitated in a translatory manner, but the longitudinal and lateral acceleration generally by means of rotation around the roll and pitch axes, using changes in angle. Of course, this cannot be described as longitudinal or lateral acceleration: It is simulated roll and pitch. Figure 2.17 shows the schematic representation of number 3. This construction is commonly found in different forms, as it offers the advantage of suggesting movement so that the driver perceives the driving environment as more realistic, but without excessive expense.

3

Fig. 2.17 Schematic representation of a dynamic driving simulator with information on the shape of the road surface and the longitudinal and lateral acceleration (dF = 3)

One early example which was put into practice in 1976 is the driving simulator at the Forschungsgesellschaft für Angewandte Naturwissenschaften (FGAN – Research Establishment for Applied Science) in Germany, which uses two cylinders in the back of the cab to replicate roll and pitch angle, superimposing vibrations, see Fig. 2.18.

2.4 Training Media

41

Fig. 2.18 FGAN driving simulator with hydraulic roll and pitch movement cues, as well as vibration (dF = 2+)

2.4.5.3 Constructions with More than Three Degrees of Freedom What was once a static driving simulator with a mock-up cab and primary controls is placed, along with a moving screen, on an electrically or hydraulically operated system (generally a hexapod), thus turning it into a dynamic simulator with six degrees of freedom (dF = 6). All three translations and the three rotations are, of course, limited, see number 4 Fig. 2.19.

4

Fig. 2.19 Schematic representation of a dynamic moving screen driving simulator with a mock-up and vestibular information (dF = 6)

One advantage of this construction is that vestibular perception is cued quickly; one disadvantage is the limited angle and path, as described above. Nonetheless, this design is very well suited for studying particularly dynamic driving tasks (partly due to its relatively low weight), if less so for regular teaching or training because of limited face validity. Below, Fig. 2.20 shows the University of Saragossa’s SYMUSYS as a rigorously well-constructed non-hexapod example. This innovative simulator

42


Fig. 2.20 SIMUSYS driving simulator from the University of Saragossa with movement (dF = 4)

with spherical platform and 4 degrees of freedom is used for entertainment and research applications with extreme driving situations. The next construction locks the out-of-window view in place as a fixed screen and uses modified bodywork close to the original. This markedly improves the face validity and sound simulation. The field of view is more realistic and also suited to perception of objects behind the driver. This offers the advantage of the vehicle being in a better position relative to the surrounding traffic, or the driver knowing the position better, depending on the quality of visibility and movement; see number 5 in Fig. 2.21.

5

Fig. 2.21 Schematic representation of a dynamic driving simulator with near-original bodywork, a fixed screen (dF = 6)

2.4 Training Media

43

There are plenty practical examples for research and training applications; one which is well-known and relatively new is the truck simulator at the University of Munich for studying traffic safety with an electric motion system at operating height in Fig. 2.22.

Fig. 2.22 Truck simulator with hydraulic movement and fixed screen, chair of automotive technology at the Technische Universität München (Technical University Munich, dF = 6)

Here, too, there are improvements. If the hexapod is placed on a platform which can be driven back and forth or sideways and the out-of-window view is also integrated on the platform, the current final stage of driving simulator development is reached, see number 6 in Fig. 2.23. This construction is known as the 6+ degree of freedom version. The moving screen and movements caused by the vehicle suspension relative to the platform create quite a realistic impression of driving. A rail system below moves the cab and the hexapod, e.g. and allows for extended acceleration duration times.

6

Fig. 2.23 Schematic representation of a dynamic driving simulator with near-original bodywork and close-to-reality vestibular information (dF = 6+)

There are few practical examples at the upper end of the scale. One example in practice is the VTI driving simulator in Sweden, partly including a car mock-up

44


shown in Fig. 2.24. This non-hexapod rail system allows for either lateral or - if the upper part including cab is revolved by 90 degrees–for longitudinal movements and is used in research applications.

Fig. 2.24 VTI driving simulator in Sweden with hydraulic movement and a moving screen (dF = 6+)

Figure 2.25 shows Daimler’s research simulator in Berlin. This is a hexapod placed on a rail system. Again, it allows for either lateral or – if the dome is revolved by 90 degrees – for longitudinal movements and is used in dynamic research applications. Longitudinal movements have not yet been realized. As well as this attempt at structuring, there are a range of other possible taxonomies for driving simulators. For example, some interesting versions with four degrees of freedom have not been mentioned at all, so as not to complicate the facts further. Of course, this categorization is technology-driven, yet it is also anthropocentric, as it is based on known needs in terms of perceptions of movement and out-of-window view, from the driver’s point of view. This overview should, however, suffice simply to illuminate the fact that not all driving simulators are alike. Also, unlike vehicles, for example, it is not at all clear which device is best suited for which use. When buying a car, a sports-loving driver would hardly consider a minivan. Categorizations of vehicle construction types are familiar to everyone as “model ranges” and are clearly defined, even though their numbers are tending to increase. When buying a driving simulator it would not be as easy to put it into a category. For a driving school or training company, high-end constructions can certainly be ruled out due to their construction costs alone. Some low-cost constructions are also poorly suited to practical driver training, as they cannot replicate aspects of driving which are fundamental, especially for teaching and training purposes, and their face validity is closer to that of simulator games. However, falling prices, new processes and higher-performing technology do not make it any less necessary to clarify the variety of requirements and technological solutions using “driving simulators.” It is

2.4 Training Media

45

Fig. 2.25 Daimler’s driving simulator with extensive movement and a moving screen (dF = 6+)

left up to discussion if the driving simulator number 5 shown in Fig. 2.21 may be located at the anchor in the middle of Fig. 2.10, the point of intersection of both gradients. Anyway, this technologically driven taxonomy remains a daring attempt because the training requirements have not even been discussed yet but should play the key role. Nevertheless taxonomies to this effect are certainly possible, bearing in mind the current state of knowledge. It seems worthwhile and necessary for the driving simulation community to identify and create these wide-scope definitions.

2.4.6 Advantages and Limits of Simulators This technologically based identification of simulator constructions leads directly to considerations of the advantages and limits of driving simulation. The general advantages of simulators have been described already early in the literature (e.g. Frieling & Sonntag, 1987; Kelley, 1969; Hoskovec, 1972; Pain et al., 1973; Pelz & Krupkat, 1974; Frieling & Sonntag, 1987; von Bressensdorf et al., 1995). It is agreed that simulators may deliver:

46


• Systemization: systematic learning strands and possibility of repeating, standardizing and automating the sequence of events and tasks; singling out subtasks • Documentation: documenting the learning and driving process with on-line error recognition and processing, measuring learning progress • Objectivization: objective checking of training process and success using parameters • Localization: creating any situation at any place, at any time; rare occurrences • Adaptation: adaptive learning to save time and individual variability of levels of difficulty, complexity • Replay: “save” function and repeating a trip for later evaluation and reflection on behavior displayed during a debriefing, see Briefings and Replays • Safety: safe learning under all conditions, critical sequences can be carried out as often as required and subject matter which cannot be practiced in reality due to danger, e.g. tire burn. Experience in aviation and ship simulation has shown, nonetheless, that there is still insufficient knowledge today about evaluation criteria for automated training evaluations and interactive training control. However, one aim of using training simulation is computer-controlled adaptation of task difficulty based on an evaluation of driving quality, meaning that a training session is shortened when the training target is achieved early. Based on the current status of our knowledge, this aim can only be achieved by means of sufficient test phases. The limits. Often, safety in simulators is only shown from its good side. However, the almost complete safety of simulators is also the main limitation of experiencing realistic driving (Biehl, 1976; Fastenmeier, 1988b; Käppler & Voß, 1988; Käppler, 1991; Käppler et al., 1992; Käppler 1993a, b, 1994a, b, 1997a). As there are no real risks in taking a decision, and no real burden of responsibility, there is no danger and the driver’s actions have no consequences. The outcome is that these aspects of real-life situations cannot be reproduced. It has been pointed out that compared with pilots on scheduled flights, for example, car drivers enjoy markedly more scope for their decisions and actions. Thus, their motivation and behavior have clear effects, and presuming a priori that people will act realistically in simulators is untenable with no further evidence (Biehl, 1976; Fastenmeier, 1988b; Käppler & Voß, 1988). Risk-free decisions and a lack of consequences can encourage an inflated view of one’s own skills and an uncritical feeling of safety. This can result in undesirable transfer, or no transfer at all, as well as uncertainty with regard to changes of attitude and behavior. It is also unlikely that misdirected attitudes and driving motivation (aggression, enjoying risk) will be unlearnt. For processes to become automatic, there must be an unbroken illusion of realistic experience. This can be disrupted if the driver experiences physical or mental discomfort, or simulator sickness. Reports of this kind have accompanied driving simulation from the very start. Findings by Förstberg (2000), e.g. showed for real vehicles use that such sickness was reported at least once by about 10 percent of train users, 9 percent of airplane passengers, 28 percent of bus riders and even 36

2.4 Training Media

47

percent of passengers in cars. Thus, in about 10–20 percent of the driving population symptoms similar to simulator sickness even appear outside simulators: this can only really be described as simulator sickness when it occurs in higher percentages in simulators. On the other hand, simulator sickness can be the result of discontinuities within one kind of information or between different kinds of information, i.e. not caused by technology. An investigation in flight simulators showed that simulator sickness occurs far more rarely when the visual display is improved, without the motion simulation being changed (Kennedy et al., 1993). Experience suggests that the same is true of driving simulators. However, corresponding technological specifications are not available for many tasks, and it is left to the skill of the developer in question to find the best solution from his or her point of view. Therefore, it is difficult to produce full-task simulators which not only provide an unbroken illusion of real experience, but also guarantee realistically risky decisions and actions with consequences. This means that the use of simulators is drawn toward a preference for cognitive subject matter and rare topics, and areas which can hardly be practiced in a vehicle at all, or can only be practiced insufficiently, for operational, political or environmental reasons. One difference compared with conventional driving school training is a result of social isolation in simulators and the effects on the learning situation itself. On one hand, the out-of-window display can only be properly perspectively adjusted to the driver’s point of view. From the passenger’s point of view, for example, distances and hidden objects seem wrong. Thus, driving instructors, as passengers, are more likely to suffer simulator sickness. An out-of-window view which was also correct for the driving instructor, for example, would be possible with virtual reality technology, but it would also be expensive, hardly acceptable for reasons of face validity and comfort, and would further restrict interaction (Käppler, 1993a, b, c). Thus, typical learning situations in driving simulators will still be unaccompanied trips without the usual interaction with driving instructors or passengers, i.e. without the conventional teacher-pupil relationship. On the other hand, the lack of other road users creates disadvantages in terms of social learning. Topics related to collective behavior require interaction with other road users. In simulators, they are usually replicated by formal models of human behavior. However, the range and variation of the corresponding algorithms are still limited, meaning that only a few situations, such as overtaking maneuvers or following the vehicle in front, are reproduced. There are often no interactive cyclists or pedestrians. Moreover, it soon becomes easy to see through the simulated road users’ limited range of reactions. Here, artificial intelligence methods and neural networks will only be able to create satisfactory results in the medium term. In the short term, multi-simulator networks offer educational opportunities of this kind by recreating social contexts realistically. Drivers interact with road users in other simulator cabs. In this case, what is problematic are learning situations where beginners interact with beginners.

48


With an awareness of their advantages, boundary conditions and disadvantages, driving simulators have been, and are still, increasingly employed in teaching and training. In the following chapters, a range of project reports are followed by two illustrative examples: • Fundamental reflections on the structure of integrated driving lessons with simulators and vehicles • Specific professional driver training courses carried out in training centers.

The end of an era . . . Safety to a . . . legislator is being able to almost demolish a concrete block. This block-busting mentality does not take in the finer points of a car which enable it to avoid hitting concrete blocks in the first place, given a driver of minimum ability. – Harvey (1985)

Chapter 3

Basic Smart Truck Driving Training Program

According to Rüter (1983), Fietkau & Timp (1989) truck drivers’ working conditions are marked by a range of characteristics. On local runs carrying hazardous materials, weekly working hours may be 60 h, mainly in the city (52 percent), 40 percent on secondary roads, 4 percent on freeways and 6 percent spent maneuvering. On long-distance runs carrying hazardous materials, weekly working hours may even rise to 70 h, with 12 percent in the city, 28 percent on secondary roads, 58 percent on freeways and only 2 percent spent maneuvering. On normal local runs, the weekly working time is 80 h; long-distance runs are characterized by long routes, night-time trips, driving in convoy, cross-border trips with customs formalities, a need for foreign languages and local knowledge. Professional drivers are subject to the following types of stress, see box:

Working Conditions of Professional Drivers (according to Rüter, 1983; Fietkau & Timp, 1989) • General lack of stimuli due to monotonous routes, routine activities, absence of people • Time pressure, time spent waiting while filling up and in holdups; bad lengths and distribution of breaks, with frequent changes between phases of activity and rest • Poor environment • High noise pollution in the vehicle and constant, monotonous rhythmic irritations (engine noises) • Gases, exhaust gases, environmental pollutants • Oscillations and vibrations • Bad visibility • Over-tiredness due to long working hours, spending too much time behind the wheel • Lack of movement, wearing gas mask


49

50

3 Basic Smart Truck Driving Training Program

• Loss of social skills, e.g. loss of ability to cooperate and communicate • Responsibility for vehicle and load (theft, acquisition, materials planning and supervision of return load) • Pressure to keep to schedule (“just in time” methods) • Frustrations (e.g. ban on passing on freeways when time is short, time spent waiting in holdups and when filling up, frequent route changes, leftover goods due to bad materials planning, gas stations too narrow to maneuver in, unfamiliar routes, or frequent checks by the police and trade association) • Physical stress when loading and unloading, connecting and disconnecting trailers or semi-trailers, and carrying and attaching the filling pipes • Nutritional problems • Inadequate sanitary and break facilities, rest and sleeping conditions • Bad visibility from the vehicle • Not enough postural changes when sitting • Risk of accident, perceived consciously or subconsciously. • Monotony, as a state of reduced mental activity, increased tiredness, diminished concentration, decreased reaction times and declining performance • Vigilance, as a state of mental alertness • Isolation, as a consequence of working and living in isolated conditions, with no co-workers and a lack of social contacts • Motivation.

As well as the hazards of driving, drivers are subject to other risks and circumstances due to schedules, short rest periods, responsibility, physical exertion, troubleshooting and maintenance, changing climatic conditions and exhaust fumes, see box. Furthermore, their motives for doing the job, such as the need for freedom, to make decisions, hold responsibility or have adventures, may not be sufficiently fulfilled. The lack of opportunities to interact with others in some aspects of the job produces social pressures. The social status of the job leads to stressful communication situations. Other stresses arise from the economic situation of many companies. Pressure from competition is passed on to the driver. For example, there are hints that drivers are expected to spend extra time behind the wheel, with reference to the difficult economic situation. This results in a conflict between safeguarding their jobs by behaving dangerously, on the one hand, and necessary safe conduct on the other. Moreover, one important stress factor mentioned by the hazardous materials drivers surveyed by Fietkau & Timp (1989) was pressure from the general public which emerged after spectacular accidents in the German towns of Herborn and Schonach: Drivers reported that they were accused of being a bunch of murderers. Long-distance drivers can suffer from a “mythical self-delusion about their own situation” (Florian, 1994), making them see their self-exploitation as a ques-

3.1 Targets

51

tion of honor. The “service of transporting goods thus turns into a heroic deed, the freeway becomes a combat arena” (ibid). Hazardous materials drivers, on the other hand, form a special subgroup among professional drivers. They are experienced, and generally educated in another profession. A special training course entitles them to operate hazardous materials motor vehicles. They are aged between 21 and 60. Stress outside the actual job includes family problems as a result of working in shifts. It has been pointed out, for example, that younger drivers’ partners are less willing to accept working hours, and to support the men. The lack of leisure opportunities and social and cultural deprivation manifests itself in increasing divorce rates. Nonetheless, or perhaps precisely for this reason, many men become truck drivers, first getting a driver’s license. The largest German truck-driver training organization, with 30,000 driver’s license applicants a year, i.e. more than 50 percent, carried out an extensive project which involved finding out how much of the training course could be carried out, at least in part, using simulators. The aim was to cut down the load on city traffic caused by holdups, pollution, wear and noise, as well as reducing the limits of the course itself: During peak times, it was hard to explain teaching matter, or city centers were even closed off to learners. As there was no available methodology and no model concept for this type of project, a methodology was developed which allowed the existing curriculum to be applied to simulators as far as possible without changes. This new method is presented in Chap. 3 and brings together the results by Käppler (1994a, b) and Käppler & Motz (1995). It is understood as an approach to simulator-assisted training and is a description of the method itself. The steps carried out are described as an example, without going into too much detail. Building upon this, in the following chapter, Chap. 4, it will be shown how the qualifications of truck drivers already working in their profession can be enhanced by simulator training courses, to improve reliability.

3.1 Targets The targets of the project were: • To shift basic practical motor vehicle driving lessons from the city to simulators or computer-assisted instruction, CAI • To achieve a driving qualification equal to the conventional one using a combination of vehicle and simulator • To reduce pollution and risks to people and materials • To cut costs • To extend training and improve traffic safety. The works were based on the driver training lessons plan for the European CE driver’s license class, for trucks with trailers (Fahrschüler-Ausbildungsordnung, 2004). Other parts of basic motor vehicle driving lessons, such as legal require-

52


ments, were not taken into account. For the sake of simplicity, it was assumed that one lesson in the driving simulator was equal to and replaced one lesson in an original vehicle (Dieterich et al., 1993). The arrangement covered the following individual tasks: • Investigating, evaluating and analyzing the teaching matter of the driver training plan • Determining the time required for each learning step • Checking how well the teaching matter could be used in a simulator • Combining the teaching matter for simulator and vehicle in a logical way • Investigating the possibilities of computer-assisted instruction • Judging how the lessons could be extended to improve traffic safety. The results concern what is required in terms of acquiring the driving simulators themselves and the staff and infrastructural changes needed. The first step was to evaluate the steps taken during driver training according to Sect. 1 of the German Ordinance on Driver Training Lessons (Fahrschüler-Ausbildungsordnung, 2004). To do so, the researchers drew upon papers by the Federal Highway Research Institute (BASt) on the classification of learning targets, published in von Bressensdorf et al. (1995). In this project, the BASt’s description was adapted to the use of simulators and can thus be used to analyze tasks, as it integrates the behavioral descriptions and sums up the targets into the following five groups.

3.1.1 Driving Skills An applicant for a driver’s license needs to learn to operate the vehicle and how to drive it in all traffic situations. The wide range of traffic situations can hardly be reproduced in driving school lessons: Accidents are rare; particular regional and seasonal features and the traffic infrastructure place limits on what the learning strand can include. Simulators, on the other hand, can replicate the many different traffic situations to occur whenever and wherever the lesson takes place (driving in fog, on black ice, with gusts of wind, downhill in the mountains or on six-lane freeways at peak times), if situations are represented realistically (von Bressensdorf et al., 1995). Simulators provide an opportunity to carry out special programs to drill driving skills, e.g. hill starts, systematically shifting down a gear to reduce speed, or gear-shifting strategies to save fuel. The question is whether all these learning processes are to take place in the simulator. It is supposed that these skills can be acquired and reinforced by adjusting self-controlled learning processes, if the driver has enough chance to practice (winter holidays in the mountains). The assumption is that training these special skills once will have hardly any lasting effect (ibid.).

3.1 Targets

53

3.1.2 Legality In driving school lessons, learners find out about traffic regulations and how to use them; they learn how to drive according to the rules. Often enough, conditions for doing so are complicated by blind corners and confusing signs, complex situations, contradictory information or expectations, illogical signage or other road users failing to adapt to the situation. These cases can also occur too rarely in driving lessons. Learning with a simulator can be used to complement this. The conditions listed above for learning driving skills (traffic situation, representation) also apply to learning behavior which follows the rules (von Bressensdorf et al., 1995).

3.1.3 Safety Consciousness Safety-conscious drivers avoid getting into dangerous situations in the first place (defensive driving style). They perceive dangers in time, are concentrated, do not stretch themselves too far and react appropriately and quickly in difficult situations. Hazard perception and reliability can be drilled systematically in a simulator. Situations where defensive driving works well can be arranged with increasing complexity, and the consequences of inappropriate reactions can be replicated and reproduced. Traffic situations can be provided involving the most common learner mistakes (driving too close, too fast around curves), as well as situations where other drivers make typical mistakes (child runs into the street, von Bressensdorf et al., 1995). After the trip, errors are analyzed in the debriefing, and they can be avoided by learning defensive behavior. It makes sense to repeat traffic situations where the driver’s behavior was not ideal. There are a range of requirements for carrying out these learning processes: • Variable levels of difficulty • Recording a session • Showing model behavior for safety-conscious, defensive driving. Furthermore, possible errors and problems must be provided in systematic form, with variable levels of difficulty. However, their mere provision does not ensure that learning processes take place at all: recording erroneous behavior does not yet provide any analysis of it. For people to learn from their mistakes, the following criteria must be fulfilled (von Bressensdorf et al., 1995): • Analysis • Explanation • Description of model behavior.

54


3.1.4 Solidarity Learner drivers must acquire social qualifications. They need to learn to get on with other road users, have understanding for those less able than themselves and show consideration for others. The anonymity and fleeting nature of contacts in road traffic, as well as the fact that inconsiderate behavior has no consequences in the broadest sense, do not support this kind of learning process. In the simulator it is possible to see your own behavior from other road users’ point of view. This creates an opportunity to learn how to see things from a different perspective. If virtual road users in simulators behave in a communicative manner, there is a chance for learning from positive role models. To initiate communicative learning processes, i.e. so that learner drivers can learn in the simulator how to act sociably in traffic, three conditions must be fulfilled (von Bressensdorf et al., 1995): • Simulators have to present driving situations in which sociable behavior can be shown and proven worthwhile (for example, a child has to run into the street after a ball). • Situations must be interactive. This results from the nature of sociable behavior, which is always interactive and follows social rules - I react in a friendly, courteous, helpful manner to others’ behavior; they, in turn, react to this behavior, and so on. Learning sociable behavior is therefore only possible as interactive learning. • The third condition is the change of perspective. Learning processes are supported if, by recording a simulated trip, it becomes possible to see the self-same traffic situation from the perspectives of the participants affected. Admittedly, this simulator arrangement only opens up the cognitive side of social behavior (trying to understand other road users’ points of view), but not the affective dimension (identifying with other road users’ emotions, empathizing with them). Empathy, a prerequisite of helpfulness and courtesy, can only be produced when dealing with people, and not when dealing with simulated road users or homunculi.

3.1.5 Morality A person driving a motor vehicle discovers that his actions are reflected with standards and values, but are also guided by these standards and values. However, pressure to act, and act quickly, rarely leaves any time for considerations of this nature. This is most likely in conflict situations. Documenting driving in a simulator, recording it and being able to quantify data (side clearance, position in lane, longitudinal clearance) means that the driving itself, driving quality, can become a subject of moral discussion and conflict analysis (von Bressensdorf et al., 1995). People learning to drive develop an image of the ideal driver and gather experience about themselves as drivers. They get to know their own cognitive processes,

3.2 Analysis of CE Driver Training Plan

55

attitudes and affects, and can learn to control their behavior. Self-reflectivity and self-control can be supported by recording on simulators, followed by reflection. For these learning processes, the simulator offers the advantage of objective, documentary recording and evaluation of a trip on the simulator, which compares favorably with driving instructors’ subjective assessments. This impartial feedback is a prerequisite for the future responsible road user to have moral, self-reflective thoughts. Still, the data only provide information on the actual behavior displayed, and not on the driver’s motives and attitudes. This, however, is the precise aim of the learning processes at this stage. These points can only be brought up in discursive analysis with the driving instructor. One example: The simulator trip records the fact that the learner driver does not cross an intersection. However, it says nothing about the learner’s motives and reasons, which can vary widely (estimate of own vehicle’s acceleration, belief that something is banned, safety considerations, uncertainty, inability, distraction, lack of concentration). Only an analysis with a skilled driving instructor can show this side of behavior, which cannot be observed, but only inferred. This analysis has a high learning effect, as it means that estimates, arguments and evaluations can be exchanged, settled and internalized (von Bressensdorf et al., 1995).

3.2 Analysis of CE Driver Training Plan After its contents page and a description of its intention, the driver training plan contains descriptions of tasks in the following 50 learning steps (see Table 3.1). The trips at the proficiency stages (nos. 45–49) are compulsory lessons which cannot be transferred to the driving simulator for legal reasons. This also applies to the driving test, step 50. Here, there have been pushes to use the opportunities simulators provide for objectifying test procedures and results (standardization, reproducibility, measurability) and to transfer the driving test, in particular, to simulators. The law does not yet allow this. This work is thus limited to the basic stages A, B and C and the practice and competency stages of the driver training plan, i.e. steps 1–44. In the face of transferring them to the simulator, all the learning steps were classified according to the following required characteristics: • • • •

Learner drivers are in the driver’s cab during the step Real vehicle not required Driving instructor not required either in or near the vehicle No guide required to give maneuvering instructions.

The available information was collected in a data sheet for task decomposition. The form used is shown in Table 3.1. In the column headings, under nos. 1–9, are the categories and individual aspects of the task decomposition for the 44 learning steps which came into question. The way the data is presented and structured was taken from Kirwan & Ainsworth (1993).

56

Table 3.1 Data sheet for task decomposition Character of task

Learning target fulfillment

Information requirement

Driver’s action

Description of the subtask from CE driver training plan

Difficulty Danger Routine Speed Precision Concentration Complexity Interactivity

Skills Legality Safety consciousness Solidarity Morality

Displays Perception Controls Cognition Out-ofMotor window Communication view Sounds Motion Communication

Feedback

Criteria for success

Displays Assessment Controls Measurement Out-ofvalues. window view Sounds Motion Communication

Staff and infrastructure

Errors and their causes

Driving instructor in/near vehicle, monitor Guide Parking lot Practice site Secondary road City Freeway

Errors of perception Errors of intention Errors of action Disruptions Organizational, human, technological influences


Task description


57

The task description (Column 1) is followed by the task character (Column 2). Column 3 shows the degree to which the learning targets are fulfilled. In Column 4, there is the information a driver needs to carry out the task, e.g. whether displays or controls must be available, as well as the communication required with other road users. Column 5 shows the driver’s reactions. These have been cut down, following Salvendy (2006), into perception, cognition, motor actions and communication. The feedback the driver receives is listed in Column 6 in a similar way to the structure of the information in Column 4. Column 7 covers criteria to assess the learning outcome (by the driving instructor: good, bad) and physical measurement values for evaluating the quality of driving and control operation. These measurement values are collected in the vehicle or driving simulator; see the outline in the following chapters. The necessary staff and infrastructure are summed up in Column 8. This relates to the driving instructor and guides giving maneuvering directions, as well as the practice site, secondary roads, etc. Column 9 contains possible errors and their causes, explained in more detail above (errors, disruptions, influence of organization, people and technology), e.g. tiredness, lack of concentration, faults (burst tires, engine damage), the effects of weather such as fog, snow, rain, and roadway conditions, such as black ice. One important criterion in assessing expenses is the time required in the simulator. Ultimately, the time needed provides information on how economical training is in a simulator. The number of actual training hours for each learning step was determined by driving instructors at a truck training center using existing planning overviews and duty rosters. The average was derived from six training groups, each with one driving instructor and four learner drivers, on the topics of introduction, mechanical inspection and familiarization with the tools; driving on private roads; basic driving skills and checking trailer safety, inspecting the vehicle before driving, etc. These figures were compared with actual values. To this effect, driving instructors received time record sheets where they recorded the number of hours spent on the lesson. These proposed times were estimated by experienced driving instructors. The margin of error for this survey is around ±10 percent. The nature and size of the sample did not allow any more precise statements to be made on the statistical dispersion of the proposals. The first result is an assessment of training opportunities in the simulator, see Table 3.2. Some learning steps require the learner to leave the vehicle. This can hardly be carried out in the simulator, as a complete original vehicle would be required, and the carefully created illusion the learner has of driving a real vehicle would be disturbed. Creating a complete mock-up of a real vehicle makes very little sense for reasons of expense and infrastructure. Learning steps which, as well as creating the cab interior, require real truck parts such as the trailer coupling, trailer or safety belts on the loading area, also cannot be carried out in the simulator. The driving instructor cannot be used in the simulator cab. As the interactive communication between the learner and a guide giving maneuvering directions does not take place in the cab,

58


Table 3.2 Learning steps of the CE driver training plan and training in the simulator Step

Simulator

1 2 3 4 5 6 7

Basic stage A Introduction, technical inspection, tools Operating the tachometer Technical preparations for driving Handling the control equipment Coordinating control movements Starting up, checking the brake system and practicing starting off Braking at a low speed

no no no no no yes yes

8 9 10 11 12

Basic stage B Driving at low speed Practicing steering Practicing changing gear Starting off on slight uphill and downhill gradients Stopping at a low speed

yes yes yes yes yes

13 14 15 16 17 18 19 20 21

Basic stage C Driving according to a guide’s directions Coupling and uncoupling vehicles Inspecting the truck before driving Maneuvering the trailer into place Checking trailer safety Driving along a series of curves Turning around Driving up to a loading platform Reversing into a parking space

no no no partly no yes no no no

22 23 24 25 26 27 28 29 30 31 32

Practice stage Lane usage Complying with traffic signs and devices Keeping correct distance Behavior toward pedestrians Speed Behavior at intersections and junctions Driving uphill and downhill Anticipatory driving technique Changing lanes, overtaking and driving past others Getting in lane and turning Driving through narrow spaces

yes yes yes yes yes yes yes yes yes partly partly

training steps involving a guide must also remain in the real vehicle. Requests the driver makes of a safety attendant or passenger, for example to see the blind spot on the right, do not require any further interaction and can be replicated in the simulator. The assessment of training opportunities in the simulator are summarized in the last column of Table 3.2. Some explanations: Step 1, introduction, mechanical inspection, tools, requires a training vehicle. Steps 2–5 (operating the tachograph, technical preparations for driving, handling the control equipment and coordinating control movements) require the driving instructor as a passenger to explain and check the individual


59

learning steps. Except for Step 3, technical preparations for driving, which must be carried out in the training vehicle (checking safety belts of people on loading area), training can also take place in the simulator, as no driving is involved. In view of the learning outcomes to be expected, the possible organizational problems and expenses, there is very little point in dividing the training between the simulator and the real vehicle. It is not until Steps 6–12 that the learner drives for the first time. These learning steps are to teach fundamental driving skills. It appears worthwhile to practice these steps in the simulator. The outcome can be assessed on the monitor; direct influence can be provided by means of tips, praise and criticism. In Step 6, starting up the engine, checking the brake system and practicing starting off, the learner asks the passenger to check the blind spot on the right. In practice, the passenger is only asked to do so when turning right and checking the blind spot. In most cases, looking in the mirror is enough. In the simulator, the “all clear” feedback must come from the simulator control station, or automatically. Step 13, driving according to a guide’s directions, is carried out in the vehicle. Steps 14, coupling and uncoupling vehicles, 15, inspecting the vehicle before driving, and 17, checking trailer safety, also require the vehicle. Step 16, maneuvering the trailer into place, can partly be carried out in the simulator. Teaching the basic principle and carrying out the first exercises require the training vehicle and the presence of the driving instructor in the truck. Following on from this, maneuvering the trailer into place can be practiced in the simulator. The assessment comes from the driving instructor or by means of measurement values, such as the side clearance. Following the same argumentation as with learning steps 6–12, it makes sense to teach Step 18, driving along a series of curves, in a simulator. Teaching Step 19, turning around, 20, driving up to a loading platform, and 21, reversing into a parking space, in a simulator are impractical as according to the German Ordinance on Driver Training (Fahrschülerausbildungsverordnung) it must be carried out concurrently with Step 13, driving according to a guide’s directions. The practice stage can be transferred to the simulator except for the following steps. In Step 31, getting in lane and turning, when turning nearside to nearside, the learner driver also needs to ensure that both drivers understand one another, e.g. by clear driving. This one-way communication can be replicated in the simulator. Steps 31 and 32, driving through narrow spaces, require guides, in Step 31 when driving onto property entrances and exits. The main part of the two steps can be shifted to the simulator in addition to training in the original vehicle. Step 34, safeguarding broken-down vehicles, must always be carried out in the training vehicle, as the learner drivers need to leave the truck during the lesson and take the prescribed safety measures. Training in built-up areas (Steps 35–44) can be shifted to the simulator, except Step 35, driving onto the road and starting off. Most of Step 35 can, just like Steps 31 and 32, be taught in the simulator. In Step 38, special traffic situations, if he does not have right of way, the learner driver is only to enter an intersection after communicating with the driver giving up his right of way. This can be carried out in

60


the simulator. The few cases of driving with a guide can be taken into account in the original vehicle. The advanced driving units, Steps 45–47, training on federal roads or secondary roads, training at dusk or in darkness and on freeways, are stipulated by law. It is not possible to transfer them to the simulator. Step 48, driving in unfamiliar places, can be carried out in the simulator. However, as learner drivers have to prove, in unfamiliar places, what they have learned so far, it makes sense to teach this in the vehicle. In Step 49, driving in test conditions, the driving test situation is replicated. In this process, the driving instructor – standing in for the examiner – is of central importance, as is the truck in which the test is later to take place. Transferring Step 50, driving test, to the simulator is currently not possible for legal reasons. The results of the expert evaluation by driving instructors are shown in Table 3.3 as how pertinent individual aspects are in percentages. The CE training program covers four out of the five learning targets. It focuses on driving skills (36 percent) and safety-conscious driving (30 percent). The learning targets of legality (14 percent) and solidarity (20 percent) are hardly taught at all, and morality not at all. Almost 60 percent of the steps were said to be hard or dangerous; all the steps required the presence of a driving instructor in some form or another. One note: Here, driving instructors are making judgments on the prospects of their own profession.

Table 3.3 Results of expert evaluation Criterion

Characteristic

Frequency (%)

Learning target covered

Driving skills Legality Safety consciousness Solidarity Morality High Medium No High: one instructor, one learner Low: one instructor, n learners No Yes No Practice site Country City High Medium No Full Task Part Task No Yes Possible No

36 14 30 20 – 59 18 23 89 11 – 20 80 52 41 43 – 2 98 79 7 14 43 41 16

Teaching matter difficult or dangerous

Driving instructor required

Additional staff required Exercise environment required

CAI required

Simulator required

Course can be extended

3.3 New Training System

61

For example, a driving instructor is only expected to be able to teach 11 percent of the subject matter to several learners at a time – areas not involving driving, such as coupling and uncoupling vehicles or inspecting the vehicle before driving. In 20 percent of all learning steps, additional staff (safety attendants, guides for maneuvering) are required. More than 40 percent of all learning stages are taught in the city. In practice, contrary to the plan, Steps 23–26, as well as 29–33, are also carried out in the city. Here, the training plan actually requires training to take place on secondary roads and federal roads. Transferring the practical driver training to CAI was only considered possible for driving according to a guide’s directions. This conservative expectation can also be explained by the fact that the instructors did not have any experience with the new technology. The prospective assessment of the practicability of simulators shows that the use of full-task simulators is considered necessary for 79 percent of the learning steps, and that 7 percent can be covered using part-task simulators and handling trainers. For 84 percent of all the learning steps, improving or extending the training course by using simulators seems possible; for more than 40 percent of the steps, it is actually expected. This assessment corroborates the positive expectations as regards to high technology seen in earlier investigations.

3.3 New Training System On these foundations, an approach was developed for a new training system for practical, integrated driving lessons with trucks and simulators, initially without any computer-assisted instruction or theoretical teaching. The concept draws together learning steps with similar content and organization into the categories of vehicle training and simulator training. These were subcategorized into training units, classing the subject matter according to the following criteria: • Content of steps requires them to be carried out in a certain order • Alternating between lessons in vehicle and simulator The learning steps must be kept in their logical order. Thus, it is not possible to teach driving through narrow spaces until the learner has practiced driving according to a guide’s directions. For reasons of motivation, and for progress checks, it seems advisable to alternate between driving in the simulator and on the road. Table 3.4 shows the new training system. Column 1 shows the training units. The vehicle and simulator blocks are each numbered consecutively. Column 2 connects the training units to the teaching steps in the CE driver training plan. Column 3 shows the exercise environment required (parking space, secondary road, etc.). Column 3 also explains whether this means the virtual world of the simulator or the real environment. In Column 4, the time required is summed up as the number of training lessons for each unit. The first training unit, Vehicle 1, contains Learning Step 1: introduction, mechanical inspection, tools. This is carried out in and around the vehicle in a parking lot.

62


Table 3.4 Rough structure of the integrated training plan in accordance with CE driver training plan Training unit

Learning step in accordance with CE driver training plan

Exercise

Time reqd (h)

Vehicle 1

1. Introduction/inspection/tools

Pkg lot

5

Vehicle 2

2. Operating the tachometer 3. Technical preparations for driving 4. Handling the control equipment 5. Coordinating control movements

Pkg lot

1

Simulator 1

6. Starting up, checking the brake system, starting off 7. Braking at a low speed 8. Driving at low speed 9. Practicing steering 10. Practicing changing gear 11. Starting off in slight uphill and downhill gradients 12. Stopping at a low speed 18. Driving along a series of curves

Simulator,

3

Vehicle 3

Additional training, Steps 6–12, 18 13. Driving according to a guide’s directions 14. Coupling and uncoupling vehicles 15. Inspecting the truxkbefore driving 16. Maneuvering the trailer into place 17. Checking trailer safety 19. Turning around 20. Driving up to a loading platform 21. Reversing into a parking space 34. Safeguarding broken-down vehicles

Exercise site

5

Simulator 2

16. Maneuvering the trailer into place 22. Lane usage 23. Complying with traffic signs, devices 24. Keeping correct distance 25. Behavior toward pedestrians 26. Speed 27. Behavior at intersections, junctions 28. Driving uphill and downhill

Simulator, exercise site Secondary road

4

Simulator 3

29. Anticipatory driving technique 30. Changing lanes, overtaking, driving past others 31. Getting in lane and turning 32. Driving through narrow spaces 33. Behavior at train crossings

Simulator, secondary road

3

Vehicle 4

45. Training on federal roads or secondary roads

Secondary road

2

Simulator 4

35. Driving onto the road and starting off 36. Priority road on a turn 37. Traffic regulation by means of traffic lights, lit signs, police 38. Special traffic situations 39. Roundabouts

Simulator, city

5

exercise site


63

Table 3.4 Continued. Training unit

Learning step in accordance with CE driver training plan

Exercise

Time reqd (h)

Vehicle 5

31. Getting in lane and turning 32. Driving through narrow spaces 35. Driving onto the road and starting off 47. Training on freeways, expressways 40. Traffic calmed areas 41. Pedestrian crossings, handicapped 42. Public transport in the process of stopping 43. School buses stopping 44. Behavior toward vehicles with special rights 45. Training on federal roads or secondary roads 46. Training at dusk or in darkness 47. Training on freeways or expressways 48. Driving in unfamiliar places 49. Driving in test conditions

Secondary road City

1

Vehicle 6 Simulator 5

Vehicle 7

Freeway Simulator, city

1 4

Secondary road

3

City, country Freeway City, country City, country

2 2 3 1

This unit is not assigned to the practical training. The Vehicle 2 training unit brings together learning steps 2–5. In this training unit, the learner learns and practices how to operate the vehicle, standing in a parking lot. Simulator training unit 1 is made up of learning steps 6–12 and 18. Here, the driving exercises are transferred from the practice site to the simulator. At the start of the Vehicle 3 training unit, the subject matter learned in Simulator unit 1 is repeated and consolidated, as the learner drivers drive the truck for the first time. This is followed by training in Steps 13–17, 19–21 and 34. The Simulator 2 training unit starts with Step 16 to reinforce maneuvering into place with a trailer, in two practice lessons. Here, the widely varying time required by individual learner drivers – as experience has shown – must be taken into account. Next, learning steps 22–28 are drilled. According to the training plan, these are to be carried out on secondary roads with light traffic, whereas Step 16 is to take place on the practice area. Simulator unit 2 is planned accordingly. The Simulator 3 training unit contains Steps 29–33. Unlike the Simulator 2 unit, these are taught on secondary roads with normal traffic. This is also why Simulator units 2 and 3 are separated. Following on from this, two of the five compulsory lessons from Step 45, teaching on federal roads and secondary roads, are slotted in as training unit Vehicle 4, so that the learner can drive in public in traffic for the first time. Steps 35–39 are assigned to the next training unit, Simulator 4. Here, learners drive in city traffic for the first time. Training unit Vehicle 5 brings together the aspects of Steps 31, 32 and 35 which require a guide and cannot be practiced in the simulator. This is followed by one of the compulsory lessons from Step 47, training on freeways or expressways, as training unit Vehicle 6. The last simulator trip, training unit Simulator 5 is

64


made up of Steps 40–44. The subject matter is behavior toward others, public transport and vehicles with special rights in city traffic. Unit Vehicle 7 contains the rest of the advanced driving lessons, plus Steps 48, driving in unfamiliar places, and 49, driving in test conditions. Furthermore, learners drive in the dark and at dusk, and are assessed to see if they are ready to take the test. The practical driver training course takes 45 practice lessons on average. The training system described here transfers 19 of these lessons to the simulator. Twentysix practice lessons continue to be carried out with the training vehicle, although this includes the five practice lessons in Step 1, introduction, mechanical inspection, tools. With reference to the practical training time in question here, 40 hours, this means that half of the established driving lessons can be transferred to the simulator. Out of the 12 practice lessons in city traffic, the training system presented here transfers 9 to the simulator, i.e. 75 percent. At this point it must be added that the above time division only applies when the effectiveness of teaching in the simulator is 100 percent. This is conditional upon the training simulator accordingly replicating all the required steps of the driver training plan, in particular the driving tasks in city traffic. This poses a considerable challenge to the capacity of simulators, and applies particularly to the number, variety, functionality and authenticity of interactive city traffic scenarios, as well as the methods for systemizing the subject matter, including the performance criteria.

3.3.1 Extensions to Training Course The assessments of how the course can be extended using simulators produced the following results, see Table 3.3. Driving skills are covered in great detail. However, driving simulators enable learning steps to be practiced systematically. By means of variation, for example, learners can practice how to act at different rail crossings. Furthermore, it provides ways of confronting learners with traffic situations which do not occur during their training. For example, the effects of the weather, time of day and seasons can be replicated with trips with heavy traffic, fog, rain or black ice. The effects of vehicle dynamics can be taught, for example different driving techniques when the truck is loaded or empty. Altogether, in this respect, the use of simulators raises the prospect of the training period being extended. The learning target of 14 percent of the steps is legality. In this respect, training can be extended by the great possible range of variety in the simulator, which enable, among other things, intersection situations to be replicated systematically, with different priority regulations and different routing. In the case of safety-conscious driving, there are ways it can be extended by providing targeted traffic situations, where the most common beginners’ mistakes come up, as well as situations in which other road users make typical mistakes, including seeing it from their point of view.


65

However, next to driving skills, even today, this training unit is the most soughtafter learning goal in driver training. To extend current training courses, learning goals and subject matter related to morally conformist behavior can be developed with driving simulators. Driving simulators with an objective recording and evaluation of a trip enable the driving teacher and the learner to discuss the course and outcome of the training trip in an educationally effective manner. Here, changing points of view is also helpful.

3.3.2 Computer-Assisted Instruction The ways computer-assisted instruction can be used in practical driver training were assessed using the Computer-Assisted Instruction (CAI) guideline. According to this guideline, the aim is to achieve standardized solutions, while striving for the same or roughly the same training targets and plans, in order to cut costs, raise training effectiveness and reduce the time taken. The choice of subject matter suited to CAI is based on the following didactic, technological and economic criteria: • With other training methods or in the original vehicle, the learning targets cannot be achieved without spending an unreasonable amount of time or money, or only if risks to staff or materials are accepted • The complexity of the subject matter can only be depicted using computerassisted methods • Costs are cut visibly compared with what would otherwise have to be spent on training. Current findings show that the teaching matter from practical driver training can be recreated using the vehicle or the simulator. There is hardly any noticeable saving in terms of staff, materials or training time. There are signs that CAI can be used instead of the practical driver training, but, to qualify this statement, CAI only seems worthwhile if new training plans are developed which blur the sharp division and categorization between driving practice and theory. Yet CAI can close the existing gap between theory lessons and practical training, and meet extensive training targets for cognitive and psychomotor skills. CAI is especially well-suited to teaching complex subject matter, tricky core knowledge and technological processes, as well as training maintenance staff, as is taught in technology lessons. For the topics of driving motives and driver attitudes, more recent works favor CAI instead of practical driving lessons. It is assumed that half of this teaching matter can be covered in this way. Thus, it seems to make sense to use CAI in areas where theory and practical lessons overlap. However, a final conclusion on the ways CAI can be used in the overall concept of basic motor vehicle driving lessons will require extensive research, investigation and development work, as no methods yet exist.

66


Mr. K. drives a car Mr. K. had learned to drive, but at first did not drive very well. “So far I’ve only learned to drive one car,” he excused himself. “But one must be able to drive two, that is, the car in front of one’s own as well. Only when one observes what the driving conditions are for the car in front and can judge the obstacles it is facing does one know how to proceed with regard to that car. – Bert Brecht (1933)

Chapter 4

Smart Driver Training Programs

This chapter draws together projects which supplement driver training with simulators as in Chap. 3. In these projects, specific one-day simulator training courses were created to improve the reliability of hazardous materials drivers and bus drivers (chapter refers to Hoffmann et al., 1994; Beelitz & Frost et al., 1995; Beelitz & Käppler et al., 1995; Frost & Käppler, 1995a, b; Käppler, 1994c, 1995a, b, c, 1996a, b, c, d, e, f, g, h, i, 1997b, c, d, e, f, 2000; Käppler & Mehl, 1997, 1999a, b; Mehl & Käppler, 1998a, b). A knowledge of the interests, constraints, demands and boundary conditions of truck drivers’ work, in line with Chap. 3, does not provide a sufficient foundation for creating practical driver training programs to improve qualifications specifically. What is needed instead are precise insights into the driver’s concrete work and its risks, drawn from actual descriptions and analyses of the job. Consequently, information about professional drivers’ workplace and jobs was collected and analyzed in projects with forwarding agencies, subcontractors and bus companies. This was then used as the basis for selecting the training targets and tasks, and for designing the training course. In this chapter, these development stages are described and accompanied by a discussion on performance criteria and analysis procedures to evaluate training results, questionnaires and notes on trainers’ qualifications. Equally, “theoretical” group lessons are introduced as a supplement and framework for the simulator course, and the contents of a sample schedule is presented.

4.1 Driving Tasks in Public Buses, Hazardous Material and Packaged Goods Transport The descriptions of the driving tasks were made by task analysts traveling as passengers on selected routes. The structure of the trip was not altered; drivers were informed before the trip began. The tasks were entered in a log without mentioning the drivers anywhere by name; no evaluations were made and the employers did not receive any information. When the trip was over, interviews were carried W.D. Käppler, Smart Driver Training Simulation, c Springer-Verlag Berlin Heidelberg 2008

67

68

4 Smart Driver Training Programs

out with the drivers, their superiors, managers, dispatchers and safety officers relating to tasks during, before and after the trip (working hours, leisure hours, pay, social environment, structure, safety culture, driver’s position in the company). All information was recorded in writing. This information was structured and classified using task decomposition; critical incidents were identified, see Frost & Käppler (1995a, b) and Käppler (1995b, 1996e, h, 1997b). The accompanied Hazardous Materials (Mineral Oil) trips took place with large forwarding agencies on dry roadways and black ice, during shifts from 3 am to 11 pm. These were the following trips: • City peddle run • Overland transport on secondary roads • Overland transport on freeways. The tractor trailers, with eight-compartment tanks, were operated by drivers aged between 40 and 60. Their tasks began with taking on the vehicle, maneuvering it into place and checking the documentation (name, number of customers and their addresses, quantity and description of fuel in tank). This was then followed by filling the tank truck at the refinery, a process characterized by long waiting periods, depending on the time of day. On the trip that followed, along freeways, secondary roads and through city traffic, deliveries had to be made to several customers. Emptying all the tanks at the gas station took the drivers 1.5 h (monitoring drainage, moving drainage tubes, opening and closing covers, operating valves, filling out forms). After trips in the city, the tank trucks were refilled at the refinery, to deliver to other customers. At the end of the ten-hour shift, the trucks were parked on the forwarding agency’s courtyard and the loading papers were given in to the dispatcher. The accompanied Packaged Goods trips took place with forwarding agencies and subcontractor companies on local runs (within a radius of 75 km/47 miles from the vehicle location) and long-distance runs, during shifts from 3 am to 4 pm, with dry roadways, black ice and snowfall. Trucks with full trailers and 7.5 t straight trucks were operated by drivers aged between 25 and 45. Before the trip began, up to two hours were taken up with checking the loading papers, route planning and loading up according to the order in which the items were to be unloaded. In doing so, the type, size and weight of the goods, the address and delivery times of the companies had to be taken into account. Delivery sometimes took place with the engine running; mostly the packaged goods were not unloaded by the customer (as was actually his duty), but by the driver, due to lack of time, although insurances do not cover damage in this case. The accompanied Public Service Bus trips took place on local public passenger transport network routes in three German cities, Berlin, Cologne and Bonn, two of which had more than a million inhabitants, and on overland trips between 11 am and 8 pm on roadways which were dry or wet from the rain, using articulated buses, regular public buses and double-decker buses. Bus drivers aged between 30 and 50 started the shift, either at the bus depot or at a bus stop. At the depot, an external checkup was carried out, involving filling out a takeover slip (route number, bus

4.1 Driving Tasks in Public Buses, Hazardous Material and Packaged Goods Transport

69

number, driver number, company ID number, date, time, visual tire check, radio, bus equipment, ramp and ramp lifting indicator, damage to bodywork, mileage, faults and damage). The driving itself was a small proportion of the tasks. The distances between bus stops were short; on average, each trip took only minutes. When the bus reached a company stop, an external check was made, faults were identified and the interior was checked for garbage or items which had been left behind.

4.1.1 Selection of Critical Situations The list below shows problem zones and critical incidents with regard to the Hazardous Materials, Packaged Goods and Bus task analyses, see Käppler (1997e) followed up by Common Problem Zones for all tasks observed. The two boxes after it lists safety-related subject matter extracted from this information for Hazardous Materials & Packaged Goods and Bus Tasks.

4.1.1.1 Hazardous Materials Problem Zones • Extremely physically demanding work at gas stations (carrying and mounting filling pipes) • When badly planned by dispatcher or due to lack of time, fuel can only be partly drained off, with risk of “slosh” effect, goods not marked as hazardous materials, warnings on goods disregarded, overloading • Hazardous materials extremely dangerous • Lack of variety during hour-long trips along freeways, hardly any communication, problems with gas station owners and car drivers.

4.1.1.2 Packaged Goods Problem Zones • Delivering packaged goods for large number of customers in short time • Multiple deliveries in the same city due to different delivery times and re-loading • Working day predetermined by dispatchers (can change plans at any time by telephone), customers, traffic volume • Loading and stacking packaged goods according to order of deliveries • Quickly loading and unloading packaged goods without forklift (not available) • Reversing without guide giving maneuvering directions • Exceeding maximum working hours and time behind wheel • No breaks, or short ones • Narrow streets in cities, few opportunities to stop, pedestrian zones, narrow yards, maneuvering and reversing without guide • Navigating in unfamiliar areas with city maps during trip.

70


4.1.1.3 Bus Problem Zones • • • • • • • • • • • • • • •

Open radio channel and high noise level in bus Few tickets bought from bus driver (bought before at machine) Few bus stop announcements (automated announcement or display in bus) Driving up to bus stops often problematic as passengers or playing children at the side of the road, or parked vehicles make it hard to approach stop Bus stops following one another in short succession and short stretches between them When turning left shortly after bus stop across several lanes, driver forced to rely on other road users’ cooperation Changing between bus lane and road made harder by cyclists on bus lane who, moving at 18 km/h (11 mph), come in front of the bus again after the bus stop Rapid progress made harder by construction sites, narrow spots, double-parked cars requiring constant braking, starting up again; cooperation needed from cars Estimating side clearances quickly and correctly Estimating longitudinal clearance when approaching bus stops Driving with the passengers in mind requires gentle starting off and braking and low lateral acceleration Systematically adjusting direction of gaze when using wing mirrors before leaving bus stop, changing lane, getting into lane, and when using interior mirror before pulling out and during certain incidents Taking care with children, old folk, the disabled, prams, full bus when driving and when a lot of people get on at stops Driving past stops when the bus is full Opening front door manually when passenger volume high, in bad weather.

4.1.1.4 Common Problem Zones • • • • •

Driver talks to himself as result of lack of communication Sleep disturbances due to shift work Long working hours Changing semi-trailers on long-distance runs Keeping to timetable.

4.1 Driving Tasks in Public Buses, Hazardous Material and Packaged Goods Transport

Selection of Training Tasks for Professional Hazardous Goods and Packaged Goods Drivers • Self-control with large number of customers in short time; multiple deliveries; different delivery times; reloading; noise level • Planning workday with dispatchers, traffic volume, shifts, time behind wheel, breaks, changing trailers, planning tours with forklifts, keeping to schedule • Correct adjustment of seats, headrests and mirrors, grip on wheel • Communication • Technical defects in terms of water temperature, faulty wiring, tire burn or overheated brake system, insufficient brake pressure • Effects of tire pressure differences, loads, slosh • Space required for trains and semi-trailers on curves and multiple curves, narrow streets and sharp corners • Direction of gaze, use of mirrors, detecting hidden elements against colorful backdrops, correct side clearance • Effects of fog, snow, rain, hail, dazzling sunlight, aquaplaning • Lack of street markings, cobblestones, street sloping to side • Heavy commuter traffic • Evasion, braking, driving past others, overtaking due to obstacles, others changing lanes • Turning, maneuvering in gas stations, construction sites, narrow spots with and without guide • Priority, turning into wider streets • Emergency braking with different loads, road surface, safety distance • Stopping in front of barriers, obstacles and other vehicles.

71

72


Selection of Training Tasks for Professional Bus Drivers • Correct adjustment of seats, headrests and mirrors, grip on wheel • Communication • Direction of gaze, use of mirrors, detecting hidden elements against colorful backdrops, view into central passenger area due to correct side clearance • Effects of fog, snow, rain, hail, dazzling sunlight, aquaplaning • Lack of street markings, cobblestones, street sloping to side • Heavy commuter traffic • Driving with passengers in mind, gentle starting off and braking, driving around curves • Evasion, braking, driving past others, overtaking due to obstacles, others changing lanes • Turning left shortly after bus stop, moving across several lanes • Priority, turning into wider streets • Emergency braking with different loads, road surface, safety distance • Stopping in front of barriers, obstacles and other vehicles • Stopping precisely at the curb when a lot of passengers are waiting • Behavior toward pedestrians, children, opening vehicle doors, at pedestrian crossing, when turning off, at bus stops, cyclists when turning, overtaking on bus lane.

4.2 Aims and Concept of the Training Course This information forms the basis for creating training programs. They shall improve traffic safety. A word about traffic safety. This is a multi-dimensional mental construct with factors which have not been fully identified. Two examples: Low failure rates in the driving test (Seidemann, 1976) show that under test conditions, drivers keep to the rules for safe behavior. Nonetheless, young drivers pay the highest insurance premiums. Conversely, when repeating the theory test, a high rate of experienced drivers fail, yet they enjoy the highest no-claims bonuses. Knowledge of traffic and its rules obviously does not always lead to safe behavior. Another example: It is undeniable that low speeds promote safe driving, but as the sole safetyrelated factor they are questionable. Adjusted for exposure, older drivers have the same accident rate as young drivers, while driving at considerably lower speeds. Here, other physiological factors (vision or reaction times) and behavioral parameters (attentiveness or experience) play a role. Little imagination is required to see that the list of factors which affect traffic safety to different degrees will be a long one. It is also unsurprising that the relationships between the parameters and the accident rates are known only to a few.

4.2 Aims and Concept of the Training Course

73

Safe behavior cannot simply be drilled in Driving qualifications, general organizational circumstances and social skills alone are hardly sufficient in traffic safety. A driver who knows about adjusting his attitude and actions to risks must also reliably put this knowledge into practice. For this reason, promoting reliability is placed in the foreground. Reliability covers risk perception and assessment, motivation and attitude toward safety, as well as qualifications in the subject, and is a key human skill. Often, situational stress is sweepingly described as the cause of unreliable behavior. It is clear that in principle, practicing specific, typical actions can combat stress of this type. However, the problem lies in practicing typical actions for every potential situation which can arise. Thus, this approach is limited to prototypical challenges which occur frequently; however, this frequency also means that behavior patterns already exist. As a result, apart from drilling familiar topics, all that often remains is to appeal to drivers’ morals. At the same time, most accidents occur in situations of relatively little complexity, which the driver has already mastered, and which he should actually also have been able to deal with at the time. This is the problem addressed in the Self-Control training program. It focuses on general, unspecific causes of error and accident and produces psychologically determined causes for the emergence of human unreliability. The list of these causes is varied and extensive, and of course not all are suitable for training. This cause of error and accident is known as fixation. Similar to visual eye fixation, i.e. maintaining the gaze in a constant direction, psychological fixation describes the state in which an individual becomes obsessed with an attachment to another human, an animal, or an inanimate object or situation. Fixation to intangibles (i.e., ideas, ideologies, etc.) can also occur. An example for fixation in traffic is the driver in a hurry and fixation towards reaching the kindergarten before closing to pick up his/her child. The obsession with “just-in-time” prevents the driver from careful attention and anticipatory reaction. Among other things, this results in delays in one’s own reactions, and inattentiveness in the form of overlooking or passing over situational information, etc. Fixation may lead to speeding, close following distances or passing yellow traffic lights. Experts assign more than half of all accidents and errors to this cause. Its characteristic feature is that the variability and range of actions, perception and decisions usually available to a subject are limited to varying extents due to a fixation on events currently occurring. In the Self-Control training program, the effects of repeated disruptions to achieving a goal are demonstrated in the form of increasing tension and irritation. In this process, the driver must be amazed by his own actions and their consequences, so that he is sufficiently sensitized for the subsequent explanation of why his errors occurred and what effect they had, i.e. the driver shall start to think about the causes and the consequences of errors. Another special aspect of the safety training course is Rare Occurrences: skills the driver once achieved and mastered, but which are not practiced in everyday life, and which therefore diminish over time (e.g. emergency braking). Knowledge is

74


thus refreshed and the question is investigated of how to act when the brakes fail. Another superordinate aspect of reliable behavior is acting in a basically Anticipatory Manner. For this reason, in the design of the routes, conditions connected with anticipatory behavior were taken into account, such as hidden objects and unseen cross-traffic. Unlike safe driving, Economical Driving, the fourth topic of the course, is a very specific driving skill. It can be practiced in depth, and evaluated relatively simply using performance criteria such as fuel consumption. The aim is to reduce fuel consumption, for example, by adjusting one’s driving style to the topography, one’s own vehicle, the environment and the traffic. This requires results, e.g. fuel consumption, from different trips under the same conditions to be comparable, i.e. a learning strand must be constructed in a highly standardized, relatively uneventful way. In summary, the general aims of the simulator training programs are: • Systemization by means of standardization and reproducibility of what is known as the “learning strand,” with specific routes, conditions and levels of difficulty related to the surroundings, traffic, vehicle and events, at any location and time; for example, in a flat place in the summer, drivers can practice driving on hills in the snow • Documentation of driving by providing specific feedback; replays, for objective proof of what drivers do, enabling them to understand and recognize their own behavior with their strengths and weaknesses • Objectivization of test to measure success of course using performance data in nearly all situations under all conditions The vocationally oriented training targets are: • Connecting cognitive acquisition of knowledge from group course with the affective, emotional experience of the selected topics on the practical driving course • Experiencing one’s own actions in crucial safety-related situations, supported and accompanied by reflections on one’s own behavior, theoretical explanations and discussions about technical and behavior-dependent relationships • Creating a basis of experience shared by all participants, as a starting point for processes of change which the course is designed to set off Based on the insights and facts gained about professional drivers’ activities, boundary conditions and ways to use the simulator, the following three one-day courses for professional hazardous material and packaged goods drivers were put together, related bus driver courses may be designed on this information: • Economical driving with packaged goods, i.e. driving adapted to external conditions with the aim of conserving resources of any type and saving fuel • Anticipatory driving technique and behavior with hazardous materials, reflecting the driver’s own core style of safe driving. Integrating the course Rare Occurrence: Brake Failure, thus turning a rare, but highly safety-relevant point into a central topic • Frustration-resistant driving and self-control, involving setting up psychologically determined error and road rage situations for drivers to experience

4.3 Economical Driving

75

Of course, other themes and combinations can also be set down, such as tire burn instead of brake failure, readaptation instead of fixation. Another aim of the course is to improve sensorimotor skills. If accidents can be traced back to the subconscious acceptance of risky conditions, stress and overtaxing, the influence of these processes is also shown, i.e. recognizing, demonstrating, reflecting on, learning safe behavior and making it automatic so that it can be adopted consistently in everyday working life. For this reason, group training and simulator-assisted training form a reciprocal relationship. The idea is for the reliability-improving subject matter taught in the group training course to be put into practice straight away by dealing with critical scenarios in the driving simulator. Conversely, on the group training course, drivers reflect upon and process experiences from the simulator. Furthermore, the simulator training enables all participants to build up a shared base of experience. Next, the three training programs mentioned are described, bringing together information from all of them which a well-informed simulator and training course operator would require to direct these courses, rather like a play, on his driving simulator stage.

4.3 Economical Driving During this course, which takes place on one day, the drivers take part in four practical driving exercises in the simulator, aimed at dealing with different tasks, see Käppler (1997c). On the introductory trip there is already a first diagnosis of basic driving skills such as the use of the correct engine speed range. Further training trips are available on various topics and can be arranged in any order depending on the driver’s qualifications, e.g.: • Anticipatory driving technique with hidden objects and when cross-traffic has priority • Driving up and down steep hills, over hilltops and down dips, using momentum and coasting phases • Choosing the correct engine speed ranges, energy-saving use of engine and foot braking, retarder etc. The corresponding driving task involves driving a fully loaded tractor trailer from a loading station to an unloading bay while saving energy as much as possible. The trip goes along a secondary road with different up- and downhill inclines to the destination. After the driver is introduced to the task and gets used to the simulator by driving along a flat road, the first, long uphill climb begins. The truck drives across a short flat area and a slight downward incline, followed by a dip and then the next steep rise. Here, the use of coasting and momentum have a decisive effect on the speed the driver can reach on the hill; the aim is to practice achieving the ideal speed when the bottom of the hill is reached, while complying with regulations and speed limits. At the highest point on the hill, there is then a rise, followed by a short downhill incline with very short visibility. Here, the driver has to choose between economical and safe driving and use his momentum to go over the brow of the hill. The idea is

76


for him to learn not to take any risks by driving too fast. On the level stretch which follows, there is a special situation: an intersection with traffic lights which change as the truck approaches, practicing anticipation, preparedness to brake and braking to reach a target. Sometimes, the light can change during the stopping process to practice the driver’s preparedness to use the remaining momentum. During the trip which follows, across a plateau, through a landscape of woods and lakes, with objects hidden around curves, the driver practices anticipation and staying at a suitable speed. The road then goes down a long slope with different inclines to a level area. Here, the aim is to practice using momentum and coasting, as well as engine and foot braking, and the retarder. There is a blind Give Way intersection with a blocked view. The driver has to judge the right momentum, watching out for cross-traffic and deciding what speed he is going to cross at. As with the intersection with traffic lights on the hills, turning off can be practiced here. After this, a final stretch heralds the end of the trip. The course tasks include: • • • • • • • • • • • • •

Introduction to the task and the simulator Training trip on a level stretch Acceleration lane to enter the road easier Flat secondary road to get used to the simulator (starting off, braking and changing gear), to the task, and to driving smoothly, as well as up- and downhill inclines, dips and rises Freeway entrances and exits Intersections with and without right of way, blind and with a clear view Stop signs or traffic lights which change as the driver approaches Continuing on a level, uphill or downhill road after stopping Coasting and being unable to see the route ahead Taking curves smoothly, even at high speeds Normal surroundings for the time of day and season, weather and road conditions Freely flowing commuter traffic without any interaction No critical events, no accidents.

4.3.1 Learning Strand and Route In the following, this description of the training task and an imaginary trip are used to isolate a description of a suitable learning strand allowing all the tasks described to be covered. On this basis, an appropriate database can be designed in the driving simulator. This learning strand is characterized by specific individual modules from the following areas: • • • • • • •

Route Surrounding conditions Traffic conditions Events Vehicles Levels of difficulty Parameters for evaluation.


77

A database for the Economic Driving course needs to include the following route, described here as an example. It connects two imaginary or existing freeway intersections, and is divided into 18 situation modules, see Table 4.1. It shows all the information prepared and required for the database designer to create this route. It runs from west to east, starting at “Western Freeway Junction.” The first column shows the brief description of the Situation module, column 2 shows the combined Total length of the situation group and column 3 the Length of the actual section module. Column 4 shows the average Trip duration in the corresponding situation module, column 5 the combined Total trip duration. Column 6 shows a short Description of the most important elements of the situation module, column 7 contains information on Fixtures. Some comments on this: The total length of the route in Table 4.1 is 16.3 km (10.13 miles). The basic direction of the route is straight ahead: it finishes in the same direction as it begins. At an average speed of 40 km/h (25 mph) the net duration of the trip between the Eastern and Western freeway intersections is about 20 min. The total elevation of the route is 0 m; the route finishes at the same level above sea level as it begins. The Uphill modules can be used as Downhill grades in the opposite direction. All curve radii are designed in such a way that they can also be driven around easily at excessive speeds and the driver’s vehicle does not tip over. Intersection 1, with the traffic lights, has a clear view. The lights change as the driver approaches, training anticipation and momentum. Intersection 2, with a crossroad which has right of way, is a blind intersection, making the cross-traffic hard to see. The cul-de-sac at the start, and Intersections 1 and 2 with cul-de-sacs, can lead nowhere and be used for putting aside the vehicle, for driving into the cul-de-sac and stopping, e.g. to shorten the session. The roadbed, road width, curve radii, transition curves, narrow spots, transverse and longitudinal slopes, changes in elevation and rises and dips all follow the German Guidelines on Road Construction (Richtlinien für die Anlage von Straßen, RAS; Forschungsgesellschaft für Straßen- und Verkehrswesen, 1984), based on a design speed of 80 km/h (50 mph). The design and position of the markings and signs follow the relevant guidelines. The whole stretch is set in fields bordered by woods on the left and right; otherwise, comparatively few fixtures are required. Other conditions can be individually adapted to the degree of difficulty the exercise requires. The surrounding conditions are: • Season preferably spring or summer • Time of day/position of sun: noon • Dry weather. Visibility must be set at a maximum. The traffic conditions are: heavy commuter traffic in the driver’s lane and the oncoming lane without obstructing the driver’s vehicle. It should not be possible for an accident to occur, which is why traffic lights only change when vehicle approaches and cross-traffic prevents the driver going across the intersections. The range of vehicles comprises straight trucks by different manufacturers, in 7.5, 12 and 24 t, which are driven fully loaded with low-power engines. This requires particular sensitivity to efficient driving. The truck fleet is complemented by various fully loaded semi-trailers and two-axle trailers.

78


Table 4.1 Economical driving situation group data base script Situation module

Total Length of Trip length section durations yds yds

Total durations

Description

Fixtures

1. Secondary road, flat

1,640

1,640

135

135

Cul-de-sac on right after 218 yds

2. Secondary 4,375 road, 5% grade

2,735

225

360

From Western Freeway Intersection, eastward Right-hand curve, R 1,094 yds


5,140

765

63

423

Straight ahead

4. Secondary road, −5% grade

5,687

547

45

468

Straight ahead

5. Secondary road, dip

5,687

–

–

468

Straight ahead

6. Secondary 8,420 road, 7% grade

2735

225

693

7. Secondary road, hilltop

8,420

–

–

693

Straight ahead Decreasing visibility


8,749

328

27

720

Straight ahead

9. Secondary road, flat 10. Secondary road, Intersection 1 11. Secondary road, flat

9,295

547

45

765

Straight ahead

–

–

765

Crossroad

Traffic lights and signs

656

54

819

Woods on right, lake on left


10,936 984

81

900


11,592 656

54

954


14,326 2,735

225

1,179

Right-hand curve, R 437 yds, 90 degrees Left-hand curve, R 328 yds, 180 degrees Right-hand curve, R 437 yds, 90 degrees Right-hand curve, R 1,094 yds

9,952

5% sign Stepped horizontal structures, e.g. wall

5% sign Stepped horizontal structures, e.g. wall

Stepped horizontal structures, e.g. wall Left-hand curve, 7% sign Stepped R 1,094 yds horizontal structures, e.g. wall Blind Summit sign Stepped horizontal structures, e.g. wall 5% sign Stepped horizontal structures, e.g. wall

Woods on left

Woods on right

7% sign and Gravel Shoulder Ahead Stepped horizontal structures, e.g. wall


79

Table 4.1 (continued) Situation module

Total Length of Trip length section durations yds yds

Total durations

Description

Fixtures


16,185 1,859

153

1,332

Left-hand curve, R 1,094 yds Gravel shoulder straight ahead

5% sign and Gravel Shoulder Stepped horizontal structures, e.g. wall

16. Secondary road, flat 17. Secondary road, Intersection 2 18. Secondary road, flat

17,279 1,094

90

1,422

Straight ahead

–

1,422

Crossroad with priority

45

1,467

Joining the Eastern Freeway Junction

–

17,826 547

Yield sign

Table 4.2 shows virtual truck examples and the maximum laden weight, engine performance and external measurements for straight trucks. Trailers and semi-trailers are empty or loaded with the maximum train weight. The center of the load’s gravity is such that the vehicles do not tip over. The trains have the following measurements: • Length 18.75 m (61 6 ) with trailer, 16.50 m (54 2 ) with semi-trailer • Width 2.55 m (8 4 ), height 4.00 m (13 1 ). In the interviews before the course begins, one thing which is established is how experienced the individual driver is. It seems sensible to confront experienced pro-

Table 4.2 Virtual vehicle data Truck and gears

t

HP

LxWxH in feet

MAN L 2000 8,163 6-gear

7,5

155

28 10 x8 2 x13 1

MAN L 2000 12,163 16-gear, double H

12

155

28 10 x8 2 x13 1

IVECO EuroCargo 120 E 18 6-gear

12

177

31 10 x8 2 x13 1

DAF FA 55.180 C 13 120 E 18 6-gear

13

181

31 10 x8 2 x13 1

IVECO EuroTech 240 E 30 16-gear double H

26

300

31 10 x8 2 x3,00

DAF FAR 95XF.380 26.293 16-gear double H

26

380

16 5 x8 2 x13 1

MAN F 2000 26.293 16-gear double H

26

290

16 5 x8 2 x13 1

DAF FAR 95XF.380 26.293 16-gear double H

26

380

36 1 x8 2 x13 1

MAN F 2000 26.293 16-gear double H

26

290

36 1 x8 2 x13 1

80


fessional drivers with sufficiently challenging training tasks to keep them motivated across the entire day of the course. On the other hand, a young person entering the profession could be overtaxed just by the same task and thus lose his motivation. As the participants are expected to have different levels of previous experience, it must be possible to create tasks on three levels (easy, medium and difficult) with identical learning targets. This applies to: • • • • • • • • • • • • •

Surrounding conditions Season Time of day Weather Road surface grip Visibility Traffic conditions Type and heaviness of surrounding traffic Other events Driver’s vehicles Performance Trailer or semi-trailer Load and center of gravity.

4.3.2 Implementation The evaluation uses the driving, consumption and activity data. For each task and route, parameters are to be selected depending on the learning target. Further details on how this is carried out and how to interpret the data evaluation are in the corresponding chapter. After the trip, the debriefing and replay are carried out using the specified parameters. The success of the training is established, as well as how the training is to continue. The theory modules fit in with the practical ones both in terms of time and content. They are arranged using a trainer manual and at Train The Trainer seminars, held regularly, taking into account the results of the continuous evaluation.

4.4 Anticipatory Driving Technique The Rare Occurrence aspect is part of the course on anticipatory driving technique. The aim is to drive safely and save resources by adjusting one’s driving style to the topography, one’s own vehicle, the environment and traffic. The basic premise of this training program is: its focus is on experiencing and dealing with situations known only from hearsay but not yet actively experienced. The driving task, similarly to the task specifications for Economic Driving, is to safely drive a tractor trailer with a full load of gasoline out of a city and through a hilly landscape. After the driver is introduced to the task and gets used to the simulator, the trip takes place up and downhill through the hilly landscape described. However, on the last section of the trip, the brakes fail, as follows. On a steep slope in a slight right-hand curve, an oncoming vehicle moves out of lane and starts

4.4 Anticipatory Driving Technique

81

to overtake. The truck driver has to brake and notices that the wheels are locking up. The trailer starts to swing out and threatens to jackknife. In this situation, the driver has to decide whether to continue braking and countersteer, or whether to swerve onto the verge or the gravel shoulder. In either case, the training session finishes without the trailer turning over or an accident occurring. If used, the stretch which follows can then act as a transition to the end of the trip and the session. Drivers learn to defuse even dangerous situations by anticipatory, timely, appropriate behavior. There are signs of the brake failure beforehand: an attentive driver would have noticed the wheels locking up earlier, or a warning lamp or unusual reaction when braking, and reacted in time. In the event, he has to decide quickly whether to move out the way or continue to brake, trusting that the brakes will be effective enough and he is skilled enough to control the vehicle. Thus, the course prepares him for a Rare Occurrence by confronting him with possible and workable reactions. The training trip is repeated with different aims, e.g. correct gear-changing, utilization of momentum, or stopping on the gravel shoulder, and the driver and trainer work on the training targets named above, developing the driver’s behavior and actions. As regards content, the anticipatory driving technique practices a kind of behavior where the driver modifies his course of action to fit expected future needs and peculiarities of a situation. One example: As a driver approaches a heavily used priority road, he reduces his own speed, using momentum, connected with information on cross-traffic, getting ready to brake, getting into the right gear, etc. First, the aim is to introduce the driver to theoretical criteria, the advantages and opportunities of using anticipatory driving. In the practical driving section, the idea is to demonstrate how well the driver uses anticipatory driving, and to improve it. This developing reflection is supported by data capture and analysis of individual behavior and actions. The evaluation is meant to show the driver the level of his own skills as well as giving him pointers for improvement.

4.4.1 Learning Strand and Implementation The learning strand corresponds to that of Economic Driving, with the modules described there: • • • • • • •

Route Surrounding conditions Traffic conditions Events Vehicles Levels of difficulty Parameters for evaluating success.

The database for Economic Driving can therefore be used for the Anticipatory Driving Technique course. The surroundings and traffic conditions, events, vehicles and parameters can be selected, depending upon the individual level of the exercise, using the descriptions for Economic Driving.

82


During the course, which takes place on one day, the drivers take part in four exercise sequences in all, see Schedule. By comparing the targets and results, specific individual problems are dealt with until the training moves into line with the parameters. In the briefing, the task is described and the parameters are presented. In the simulator, the matter taught is then put into practice, and finally the result is analyzed. On the introductory trip there is already a first diagnosis of basic driving skills such as anticipatory gear changes. If necessary, remedial tips are provided and further instructions given. Depending on the driver’s qualifications, further training trips on various topics can be arranged in any order, e.g.: • Anticipatory behavior with hidden objects and when cross-traffic has priority • Driving up and down steep hills, over hilltops and down dips, using momentum and coasting phases • Choosing the correct engine speed ranges, energy-saving use of engine and foot braking, retarder etc. After the trip, the debriefing and replay are carried out. In the debriefing, the driver goes into the technical details of the Rare Occurrence brake malfunction, explains the first signs and the proper countermeasures. The aim of this theoretical, reflective follow-up activity is to extend the driver’s knowledge of: • Technological functions • Possible malfunctions • Appropriate reactions to these malfunctions. The success of the course is established, as well as how the course is to continue. In the briefing for the task which follows, the next subtask is practiced, e.g. suitable reactions to the brake malfunction. The theory modules in this respect fit in with the practical modules, both in terms of time and content, see schedule. For the briefing and the debriefing, see the comments in the corresponding chapter.

4.5 Frustration-Resistant Driving and Self-Control Here, too, group training was combined with simulator-assisted training. By dealing with the disruption scenarios in the simulator, the aim is for the driver to put into practice the points taught in the group training about improving reliability. Conversely, the group training enables drivers to process the experiences gained in the simulation, see Käppler (2000). The main aim is for the driver to realize that his behavior is of central importance and that neither alcohol nor personality factors play a special role here. The course runs as follows. First, the topic of Causes of Accident is dealt with. In the practical driving section, barriers are repeatedly placed in the drivers’ way, e.g. when transporting oil, gasoline or petrol through a city, preventing them from completing the task, such as vehicles in front of them apparently braking randomly. These barriers must appear repeatedly in order for fixation to occur. It is important that the driver is highly motivated to achieve the goal. If the driver then repeatedly has to interrupt what he is doing to achieve his aim, he will become irritated and fixation, as described earlier, will occur. This is measured by means of braking reaction times,

4.5 Frustration-Resistant Driving and Self-Control

83

the time between traffic lights change or the vehicle in front brakes and the trainees own braking. In the theory section, this psychological context is examined in more detail. The aim is for the driver to develop a picture of why accidents can occur, and to discover how he too can succumb to specific errors despite himself. He will then be able to recognize the error causes and when and why these errors are developing and counteract this in everyday life. The following program is carried out: During the introductory trip, a test event is integrated into a normal, unstressful trip to compare driver reaction times. The parameter for these comparisons is reaction time, e.g. when braking because of the following reasons, rather than a possible accident: • Abrupt braking maneuver by a vehicle ahead • Children running onto the street suddenly. The point is to show how drivers’ reaction times plummet due to the circumstances described. In the next training sessions, barriers or disruptions are placed in the driver’s path in the form of various realistic, environmentally valid situations. Some examples are: • Traffic lights constantly require driver to stop • Vehicle ahead accelerates and slows down erratically but cannot be overtaken. Experience in aviation and shipping has shown that six such interruptions in one training session are enough to irritate drivers, making them increasingly fixate on the event currently unfolding and start to overlook important events. At the same time, the test event is inserted, such as the child running into the street. The drivers’ braking reaction times grow measurably slower and dangerous for the virtual child. The aim of the Self-Control course is to demonstrate this phenomenon to drivers, tangibly, by means of their own actions. In the accompanying theory blocks, drivers are told how this effect comes about and how to deal with it. The central approach of this preventive measure is experiencing and reflecting on conditions required to improve reliability.

4.5.1 Learning Strand, Disruption Scenarios and Situational Events The driving task is to drive a truck safely through a city, along a secondary road or freeway: to fulfill this task, the scenarios in the city, on secondary streets or freeways which are already in the simulator, can doubtless be used with no major restrictions, as long as they fulfill the standard minimum requirements for surrounding conditions, traffic conditions, range of vehicles and measurement values. It is not necessary to set up a specific database, in particular a route as in the Anticipatory Driving Technique course. For this reason, no specific explanations are provided here about the content of the training task or the sequence of events. Details are shown in Käppler (2000). However, the course only comes to life when a blend of situations adapted to the training purpose are included; these can be integrated in modular form. As the success of the Self-Control course depends on the design and realism of these events,

84


there now follows a detailed description of which events are used, and why, with all the conditions required for employing them. It goes without saying that not all situations must be carried out, but only a selection. The following types of event are absolutely necessary: • Disruptions. It must be possible to integrate disruptions into the course of the trip on all routes. • Situational events. Test situations where the driver’s reactions to disruptions are measured (reaction times, speed, braking, etc.). Disruptions muddle up the normal course of the trip. They create stress, irritation, frustration and aggression, in order to measure the reaction times in Situational Events. The following aspects are crucial for their deployment: • • • •

Driver must act in motivated, target-oriented manner Events must be well placed and timed Events must be unexpected, no familiarization effects Tests must be possible during the Situational Events.

So that the disruptions do not appear arbitrary, most of the trip must remain normal, i.e. the events must appear as additions to a goal-oriented task. Here, it is left to the trainer to judge the type and number of disruptions to be used. The trainer will not prepare the driver for the events, and not repeat the same situations frequently, or overtax the driver. It must be possible to choose the time and place the events occur, in most cases. The following tables, Tables 4.3, 4.4, 4.5, 4.6, describe possible disruptions caused by: • • • •

Other road users Technical malfunctions Road layout The weather.

The first column of the tables contains the disruption, followed by a short description in the second column and adjustable variables to vary the disruptions in the third column, e.g. the distance between the vehicles and their speed. OV means the driver’s own vehicle; ORU means the other road user who sets off the event. Basically, the variables of traffic volume and visibility can be adjusted to change the level of difficulty and for variation with nearly all the following disruptions, but in Tables 4.3, 4.4, 4.5, 4.6 this is not mentioned separately, for the sake of clarity. Table 4.3 shows Disruptions Caused by Other Road Users. The table also does not include the following variables: • General weather and road conditions • Type of road user causing the event (car, truck, bus, motorcycle, cyclist, pedestrian, child, old person, inebriated person). Table 4.4 Disruptions Caused by Technical Malfunction of Own Vehicle comprises events such as brakes not working effectively, engine overheating or tire blowouts, which can occur in any traffic situation. Here, the time and duration of the disruptions can generally be adjusted; this is not listed separately in the corresponding table.


85

Table 4.3 Disruptions caused by other road users (ORU), own vehicle (OV) Disruption

Description

Variable

Failure to give way Cutting in Braking

OV’s priority ignored by ORUs ORU changes into OV’s lane directly in front of OV ORU brakes hard and unexpectedly in front of OV

Aggressive tailgating ORU leaves pull-off

ORU tailgates OV and constantly swings out, flashing headlights and blinkers As OV approaches, ORU leaves pull-off close in front of OV and filters into the traffic flow (if no pull-off, ORU blocks lane and forces OV to stop) Children cross road in front of ORU or behind ORU. When stopping ORU sets hazard warning signal flashers (additional option to ORU Leaves Pull-off ) OV blocked ORU parks on own lane As OV drives past, door of parked ORU opens suddenly on side facing the road (additional option to ORU Blocks Own Lane) Jam on road in OV’s direction Special-purpose ORU activated in traffic jam (additional option to Traffic Jam On Road) Stop-and-go traffic in traffic jam

Distance ORU–OV Distance ORU–OV Distance ORU–OV Frequency of ORU’s braking Distance ORU–OV

Children crossing

ORU blocks lane Door opening

Traffic jam Special-purpose ORU Stop-and-go traffic Oncoming ORU swings out Overtaken ORU swings out Slow ORU in front ORU leaves road Filtering into road Left turn Overtaking Cyclist lane change Cyclist turning

ORU crosses road 4-way stop

OV dazzled

Oncoming ORU swings out into OV’s lane ORU forces OV to overtake; as OV starts overtaking, ORU swings out itself Slow-moving ORU gets in OV’s way and stops at will ORU in front fishtails, slows down and leaves the road Heavy traffic makes it hard for OV to filter in Heavy oncoming traffic makes it hard for OV to turn off left due to ORU’s uncooperative behavior Overtaking ORU pulls in front of OV and gets in the way Cyclist going along right side of lane unexpectedly moves left onto lane in front of OV Before turning right, OV has to overtake cyclist; OV has to wait for pedestrians and cyclist rides into blind spot ODs cross own lane slowly in front of OV As OV approaches intersection where vehicles have equal right of way. ORU on right has right of way, other vehicles approach the intersection and it is not clear which has priority: as OV starts off, ORU moves in OV dazzled by oncoming traffic and low visibility at night, ORU and obstacles visible at last minute

Distance ORU–OV

Distance ORU–OV

Distance ORU–OV Distance ORU–OV Time ORU opens door Duration Speed of ORU Crawling speed Duration Distance ORU–OV Speed of ORU Distance ORU–OV Speed of ORU How ORU stops Speed of ORU Speed of ORU Distance ORU–OV Speed of ORU Distance ORU–OV Speed of ORU Distance ORU–OV Distance ORU–OV

Distance ORU–OV Speed of ORU

Distance ORU–OV Speed of ORU

86


Table 4.4 Disruptions caused by technical malfunctions of own vehicle OV Disruption

Description

Oil pressure warning Brake pressure warning Brake failure due to pressure

Oil pressure warning lamp lights up during trip

Brake circuit failure

Warning lamp lights up during trip; display shows it is below minimum pressure Warning lamp lights up during trip: drop in pressure due to ripped v-belt, truck can only be stopped using hand brake With no previous information, brake circuit fails so braking is only possible using: • tractor • trailer or semi-trailer brake

Brake valves stuck ABS failure Coolant temperature warning Running noises Burst tire Low tire pressure Load slips Semi-trailer overbrakes Brakes overheat

Truck can only be stopped using hand brake. During trip, warning lamp lights up, trip continues without ABS During trip, warning lamp lights up. Display shows temperature above normal range Engine and gears make noise during trip Tire suddenly loses air and starts to burn; smoke forms Tires lose air and vehicle fishtails Load slips during trip Tractor trailer overbrakes, semi-trailer swings out due to poor brake distribution Smoke builds up in the rear-view mirror; very poor braking effect

Variable

Drop in pressure Drop in pressure

Drop in pressure Tractor or trailer failure

Axle Drop in pressure

Wheel Wheel Direction of slip Speed Swinging out Braking effect

Table 4.5 shows Disruptions Caused by the Road Layout, which are set in the fixed database, e.g. construction sites with lane width limits, grade crossings or missing road markings. For this reason, only few variations can be made when running the course. Table 4.6 shows Disruptions Caused by the Weather, e.g. rain, fog or side winds, lasting across long sections of the trip. Mostly, the conditions last the whole trip long and are already set down as the starting conditions. For events in this category related to the current topic, only short-lived ones are of interest, such as slippery roads due to leaves or ice, which only affect a few meters of the road. It is difficult to create some disruptions in a realistic, acceptable manner. An attentive driver will likely master them easily; only one who is not concentrating will succumb to errors. However, the experience will be more rewarding and its effects better transferred into everyday life if the course lives up to the participants’ expectations and extends existing skills. For this reason, to extend the learning outcome beyond the attractiveness of the course, individual disruptions are put together to form complex disruption scenarios in a hierarchical structure. Among other things, this is understood as interweaving disruption scenarios, in a hierarchical structure with one building up from the next, from the areas named above: Other Road Users,


87

Table 4.5 Disruptions caused by the road layout: OV – own vehicle, ORU – other road user setting off the event Disruption

Description

Ruts

Ruts cause OV to fishtail and force driver to slow down Transverse ruts cause OV to pitch and force driver to slow down Own lane width limited, obstructing OV and forcing it to drive slowly and watch out for oncoming traffic Missing road markings combined with signage in accordance with traffic regulations makes it hard to stay in lane Freeway exit has curve which grows considerably sharper Truck train has to mount sidewalk slowly Grade crossing forces driver to go slowly Closing gates force driver to stop

Transverse ruts Lane width limited

Missing road markings Tractrix Sharp curve Grade crossing Grade crossing with gates Changing signal lights

Signal lights change depending on position and speed of OV • If no reaction, driver would cross on red • Set not working, orientation required

Variable

Construction site, vegetation Routing

Switches over depending on speed and position of OV Flashing lights on/off

Table 4.6 Disruptions caused by the weather: OV – own vehicle, ORU – other road user setting off the event Disruption

Description

Variable

Rain

Poor visibility and slippery road surface force driver to go slowly Poor visibility and slippery road surface force driver to go slowly Poor visibility and slippery road surface, brakes have no effect, fishtailing Road surface slippery with leaves, hard to see

Visibility

Snowfall Aquaplaning Slippery leaves Mist Fog Night, dusk Dazzling sun low on horizon Slippery snow Slippery ice Side winds

Poor visibility, vehicles and obstacles visible at the last minute Poor visibility, vehicles and obstacles visible at last minute Restricted visibility, vehicles and obstacles visible at last minute, forcing driver to go slowly Restricted visibility, vehicles and obstacles visible at the last minute Slippery road surface forces driver to go slowly Very slippery road surface forces driver to go slowly Difficult to stay in lane due to lateral forces

Visibility Depth of water Length of time, grip Visibility Length of time, wall of fog Time of day Darkness

Duration Duration Length (gusts)

88


Technical Malfunctions, Road Layout and Weather. For example, the disruption can be purposefully brought to a pitch by: • • • •

Increased traffic volume Low-visibility weather Distracting news playing on the radio Braking maneuvers by other road users.

This can be illustrated by an example already described in another context, and extended and changed here: An experienced professional driver has the task of driving his vehicle safely, in a manner adapted to the topography and environmental and traffic conditions. He is to drive a tanker fully loaded with premium gasoline from the start to the end location. The route passes along the hilly route described in Rare Occurrences, a secondary road with slopes. The following disruptions have been preset and make the learning strand difficult: • • • • •

Early in the morning Dark Rain Heavy commuter traffic Up- and downhill slopes.

In this scenario, the following disruptions occur one after another, spread across the trip: • • • •

Lights change from green to yellow, so the driver has to brake hard A vehicle stopping at the side of the road means he has to stop too A slow vehicle unexpectedly swings out in front of the driver An impatient driver pulls out from oncoming traffic into the other lane to overtake, shortly in front of the driver.

Initially, the disruption scenarios hamper the range of actions the driver can carry out, provoking him to react as desired. When the course is run, there need to be smooth, continuous transitions between the disruptions: the criteria relating to this is that the driver must accept the scenarios. Accordingly, the following catalogs of disruptions are to be understood as suggestions. Altogether, one of the real challenges to the course designer is to put together different kinds of disruptions from the areas described, e.g. combining bad weather, heavy traffic, a narrow roadway, mechanical brake failures and other road users swinging out, to form an accepted disruption scenario. Here, creative design is just as essential as managing to coordinate and time the individual disruptions in relation with one another.

4.5.2 Implementation Situational events are test situations. They force the driver to react and serve to measure aspects such as his reaction times when braking. The aim is to show that in the


89

Table 4.7 Test situations: OV – own vehicle, ORU – other road user setting off the event Test situation

Description

Variable

Parameters

Traffic light change

Unexpected traffic light change from green to yellow/red as OV approaches, forcing driver to stop ORU in front brakes hard and unexpectedly in front of OV, forcing driver to stop Oncoming ORU swings out suddenly into OV’s lane and forces driver to stop

Time of change can be adapted to speed of OV

Braking, gears, gas Steering wheel angle Traffic light change

Distance of ORU can be adapted to speed of OV

Braking, gears, gas Steering wheel angle When brake lights lit

Distance of ORU can be adapted to speed of OV

As OV drives past, door of parked car opens suddenly on side facing the road, forcing driver to stop Children crossing force driver to stop

Time door opens can be adapted to speed of OV

Braking, gears, gas Steering wheel angle Boundary of road crossed Braking, gears, gas Steering wheel angle Door opening

Braking vehicle

Vehicle swing out

Door opening

Children crossing

• behind bus at bus stop • between parked vehicles

Time they appear can be adapted to speed of OV

Braking, gears, gas Steering wheel angle Appearance of children

case of fixation, reaction times grow longer and accident-prone behavior increases. Table 4.7, Test Situations, describes events from the tables above, with possible tests added. The first column of the table names the Test situation, followed by a brief Description in the second column and, in the third column, adjustable independent Variable for varying the disruptions, e.g. the distance and speed of the vehicles. In the fourth column, the minimum Parameters are listed which need to be registered as time functions to measure the driver reaction times. Driver reaction times must be registered highly precisely, e.g. at 1/100 s. In the data analysis, these time functions are used to determine how much time passes between a light changing from green to yellow and the time the driver first touches the brake pedal, for example. This time is the desired parameter: the reaction time. In this table, once again, OV means own vehicle, ORU means other road users setting off the event. When putting together the disruption scenarios, the following are set down: • • • •

Surrounding conditions Traffic conditions Range of vehicles Levels of difficulty.

90


The drivers’ own vehicles can be selected: they are only restricted in terms of the level of difficulty (see below). In this course, again, accidents should not be possible: if necessary, the session should be broken off. Due to the participants’ varying previous experience, the courseware for this section must also allow similar tasks to be created with the same aims but different levels of difficulty; see the details on Economic Driving. The main evaluation parameters have been noted in the tables. As well as this, of course, it makes sense to take into account other parameters such as fuel consumption or staying in lane. Please refer to the description of the parameters in the chapter Anticipatory Driving Technique. In a closed-loop feedback system with the trainer and a comparison with target and actual results, the driver learns that every individual disruption in itself (slopes, rain, commuter traffic, full load, other drivers, etc) requires his attention. He will only succeed in managing these disruptions and irritating situations if he learns to assess his situation properly, to notice when fixation is increasing from event to event, and to defuse the situation by making decisions and acting well in time. In fine-honed briefings before the trip, a specific topic is presented, as well as the task and parameters. The task is then carried out and analyzed in the simulator, followed by an evaluation and further instructions in the debriefing, then the briefing for the next task, and so on. A short description of the learning process. It starts off with an insight. The driver mentally processes the riskiness of his customary behavior. The aim is for him to experience in practice, for example, that driving down a slope safely means entering it at the right speed to start off with. The participants are meant to become aware of the danger they can get into if they are inattentive. After the driver is familiarized with the task and problem on the first trip, on the next trips the driver learns to: • • • •

Observe matters attentively Assess situations right Make quick, accurate decisions Act in time.

To deal with this issue, lessons are provided in the simulator as well as on truck trips. For example, the driver learns targeted and emergency braking on the training ground. In the simulator sequences, he adapts his speed to the weather conditions and time of day. From secondary signals, the driver will recognize when rushed drivers in oncoming traffic are planning to overtake. He learns that it is better to stop in time when necessary and avoid risks. Throughout, individual, practical driving exercises are practiced until they sink in; more complex ones, however, act as a deterrent and only serve to demonstrate how difficult it is to deal with a series of disruptions. On this course, the follow-up activities (debriefing and replay) are of especial importance as alternative actions can be demonstrated in a convincing manner. The result of this training unit is that the driver is sensitized to sequences and chains of individual situations turning into a disruption scenario which restricts the actions he

4.6 Evaluation of Training Results

91

can take, to a critical extent. The aim is for the driver to realize when the point of no return has been exceeded. He learns to drive attentively and understands that acting quickly is absolutely necessary to avoid accidents.

4.6 Evaluation of Training Results One advantage of simulators is that evaluations can be objective. This objectivity depends on the validity of the physical measurements used as parameters. In the following, current measurements and a method are presented, which allow the following assessments to be carried out: • Evaluation of course in progress • Evaluation of course results • Evaluation of course itself. More details are given in Käppler (1997d, i). The course currently in progress is tested during and directly after the trip using dependent primary variables. This is required to help the trainer identify problem zones and necessary interventions, such as replays, and serves as feedback for the driver directly after a trip. Secondary variables are calculated from primary variables as the basis for evaluation, allowing statements to be made on: • Learning progress • Level of achievement. The learning progress shows a driver’s current skill compared with his original ability. The absolute level of achievement is the quality of his driving compared with the stored data of an ideal driver. This normalization means that statements can be made about the driver’s absolute level of achievement, as well as direct comparisons between drivers. For these assessments, training conditions must be observed which have a standardizing effect; in other words, the trips must be subject to roughly the same conditions. Restrictions are selected specific to each exercise, e.g. a speed limit of 30, 40 or 50 on left-hand curves. In order for the driver to obey the restrictions, there are acoustic warnings; for example, when he exceeds speed restrictions or touches the curb. Experience shows that presenting the raw data is even helpful when elucidating and explaining deficient actions and their alternatives. During the simulator trip, two different classes of primary variables are recorded in the form of time functions: • Independent primary variables • Dependent primary variables. Independent primary variables show the set conditions for the trip before it began, e.g. the name and age of the driver, the type, load and parameters of his vehicle, and the events, all adjusted by the trainer before or during the course. Dependent primary variables are measurements of the driver’s reactions to these conditions. They result

92


Table 4.8 Independent primary variables Primary variables Name of courseware Day of course Name of trainer Name of participant Roadway, road width Parameters of driver’s vehicle: Output, gears Yaw amplification Tires Cut-off frequencies Transport delay Roll amplification Length, width, height Mass, payload Stability factor Steering ratio Wheel base Event around driver’s vehicle Traffic lights Traffic signs Pedestrians Vehicles

Comments Date and time

Pursuant to German Guidelines on Road Construction These data are input parameters for the vehicle model and must be taken from this model

Name Time, status it occurs/appears Position (x, y), distance in radius of 250 m (273 yds) Angle of yaw, speed Status of brake lights and blinkers

from a trip and are stored in the simulation computer as time functions, e.g. the lateral deviation, driving speed, fuel consumption, etc. Table 4.8 shows Independent primary variables in the first column and comments in the second. Table 4.9 shows the Dependent variables for evaluating course in progress. In Column 1, it shows on-line control variables, which are to be shown on the simulator monitor as raw data over time. Column 2 shows comments, Column 3 the unit of measurement. The courseware must allow users to select and display these control variables in different ways. After the trip, the variables selected are processed statistically, e.g. as mean values, standard deviations, minimum and maximum values, and issued for the debriefing. The secondary variables are analyzed statistically and presented as raw data, and to show the level of achievement and learning progress. As well as the raw data on times, it also makes sense to provide mean values and standard deviations, minimum and maximum values, e.g. for the side clearance, longitudinal clearance and driving speed. To assess the level of achievement, secondary variables are compared with stored ideal values, indicated as a percentage and weighted. Different weighting factors are proposed for the results of two different subtask classes:


93

Table 4.9 Dependent variables for evaluating course in progress On-line control variables

Comment

Unit of measurement

Driving time Idle time Route already covered Longitudinal clearance in front Side clearance to right

Since start of exercise Since start of exercise Since start of exercise Distance to next traffic element Distance to traffic elements to the side Distance to traffic elements to the side

s s m (yds) m (yds)

Side clearance to left Driving speed Gear changes Fuel consumption

Number

m (yds) m (yds) m/s (yds/s) n l/100 km (gal/mile)

• For uninterrupted trips: 0.66 (driving straight ahead and around curves) • For situational and disruptive events: 0.33 (emergency stops and evasive maneuvers). The learning progress is measured in percentage of the driver’s own initial values, e.g. for: • • • •

Keeping to the speed limit, for restrictions to 50 km/h (30 mph) Longitudinal clearance compared to safe clearance with a reaction time of 1.5 s Braking reaction times ˙ Side clearance of 1 m unrelated to driving speed, over 6 km/h (4mph).

Table 4.10 lists Dependent primary variables for further processing into secondary data. It distinguishes between primary variables for driving quality and for operating quality. The first column shows the dependent Primary variables, the second the Unit of measurement. Table 4.11 shows the Dependent secondary variables to be processed to evaluate the course results, divided into driving quality and operating quality. The independent primary variables are simply taken from those recorded in Table 4.8. The first column shows the Secondary variables, the second Comments; Column 3 shows the Unit of measurement, Column 4 the Ideal values to be achieved, if known, and Column 5 shows the analysis and Display of the data with these typical values: • • • • •

Current raw data across time Total value Integral Mean value (MV) and standard deviation (STD) Minimal value (min) and maximum value (max).

Some comments on this: As well as the driving and idle time and the lateral deviation, the side and longitudinal clearances are also determined between the driver’s vehicle and all vehicles and traffic elements driving in front and behind. The side clearance in particular has been shown to be a good indicator of driving quality.

94


Table 4.10 Dependent primary variables for evaluating course results Primary variables

Unit of measurement

Driving time since start of exercise Idle time since start of exercise Position of driver’s vehicle (x, y, z) Longitudinal deviation since start Lateral position on lane Clearance behind to next traffic element Clearance in front, as above Side clearance to right, as above Side clearance to left, as above Longitudinal speed in direction of movement Lateral speed at right angles to this Current driving speed Longitudinal acceleration Lateral acceleration Steering wheel angle Steering wheel angle speed Steering angle acceleration Roll angle Pitch angle Yaw angle Yaw angle speed Yaw acceleration Drift angle Engine speed Brake temperature Fuel consumption Tire wear Gear wear Brake lining wear Gear Brake pedal position Gas pedal position Gear pedal position Engine brake Retarder ABS Cruise control Ignition key Light Blinker Horn Belt emergency brake Seat

s s m (yds) m (yds) m (yds) m (yds) m (yds) m (yds) m (yds) m (yds)/s m (yds)/s m (yds)/s m (yds)/s2 m (yds)/s2 degrees degrees/s degrees/s2 degrees degrees degrees degrees/s degrees/s2 degrees n/s degrees l mm (in) mm (in) mm (in) n % % % 0/1 0/1 0/1 0/1 0/1 0/1 0/1 0/1 0/1 0/1 correct


95

Table 4.11 Dependent secondary variables for course results evaluation Secondary variable Driving quality Driving time Idle time Longitudinal deviation Lateral position Lateral deviation Lateral offset Longitudinal clearances Side clearances Event near driver’s vehicle in a radius of 250 m (273 yds) Traffic light near driver’s vehicle in a radius of 250 m (273 yds) Traffic sign near driver’s vehicle in a radius of 250 m (273 yds) Pedestrians and vehicles near driver’s vehicle in a radius of 250 m (273 yds)

Safety clearance Area used Longitudinal speed Lateral speed Driving speed

Comment

Unit of measurement Ideal value

Display

since start of exercise since start of exercise Route covered since start of exercise Position of vehicle in own lane Deviation from own lane Deviation from center of lane Distance to all traffic elements in front Distance to all traffic elements to the side Description Time and status of occurrence Distance to driver’s vehicle Description Time and status of occurrence Distance to driver’s vehicle Description Time and status of occurrence Distance to driver’s vehicle Description Position (x/y) Distance to driver’s vehicle Yaw angle Speed Status of brake lights Status of brake lights Time of appearance

s s m (yds)

short

Total Total Total

m (yds)

far to the right

MV, STD

m (yds)

low

STD

m (yds)

more to right

MV

m (yds)

large

m (yds)

large

Name s, N m (yds)

none none large

current, MV min, max, MV, STD current current

Name s, N m (yds)

none none large

current current

Name s, N m (yds)

none none large

current current

Name m (yds), m (yds) m (yds) degrees m (yds)/s 0/1 0/1 s

none large large small low 1 1 early

current current current current current current current

Average distance to vehicles ahead Total area used outside of lane in the vector direction in the vector direction Current speed

m (yds)

large

m2

low

min, max, MV, STD Integral

m (yds)/s m (yds)/s m (yds)/s

as required very low according to regulations

current current current

96


Table 4.11 (continued) Secondary variable


Display

Average speed since start of trip Longitudinal negative when delayed acceleration Lateral acceleration from vehicle model

m (yds)/s m (yds)/s2

even low

m (yds)/s2

low

Steering wheel angle according to sensor

degrees

Steering wheel angle speed Steering angle acceleration Reversal steering angle Roll angle

according to analysis

degrees/s

low

MV, STD current, max, MV current, max, MV current, max, MV MV


degrees/s2

low

max, MV


n

few

Total

from vehicle model

degrees

low

Pitch angle

from vehicle model

degrees

low

Yaw angle

from vehicle model

degrees

Yaw angle speed

from vehicle model

degrees/ss

low

Yaw angle acceleration Drift angle

from vehicle model

degrees/s2

low

from vehicle model

degrees

low

current, max, MV current, max, MV current, max, MV current, max, MV current, max, MV MV, STD

Time to Line Crossing, TLC

s

long

MV, STD

Time to Collision, TTC Estimated level of task difficulty Engine speed

Time available before boundary and line crossed Time available before collision by driver on questionnaire from vehicle model

s

long

MV, STD

n

low

MV, STD

green range

MV, STD

Brake temperature

due to brake fade

Braking reaction time

Time to push pedal half-way through

Current fuel consumption Current fuel consumption Total fuel consumption Consumption when starting off

Comment

l (gal)/s degrees

low

MV, STD

s

short

current, MV, STD

Normalized consumption Total consumption

l /100 km (gal/mile) l (gal)

low

current

low

Total


l/100 km (gal/mile)

low

MV

for 0–60, 0–80 km/h (0–37, 0–50 mph) final speed

I

low

MV


97

Table 4.11 (continued) Secondary variable

Comment


Consumption across selected for certain l/100 km route upward and downward (gal/mile) slopes Tonnage Consumption per metric l/100 km consumption ton of payload (gal/mile)/t Operating quality

Display

low

MV

low

MV

Total

Gear

Number of gear changes n

few

Brake pedal use

Number of uses

l (gal)/s

few

Total

Brake pedal speed

Type of use

l (gal)/s

low

MV, STD

Gas pedal speed

Type of use

l (gal)/s

slow

MV, STD

Gear pedal path

Number of uses

%

few

Total

Driving brake

Number of uses

n

rare

Total

Engine brake

Number of uses

n

frequent

Total

Retarder

Number of uses

n

frequent

Total

Cruise control

Number of uses

n

frequent

Total

Ignition key

Use

y/n

once

Total

Light

Use

y/n

when needed

Total

Blinker

Use

y/n

when needed

Total

Horn

Use

y/n

rare

Total

Belt

Use

y/n

always

Total

emergency brake

Use

y/n

frequent

Total

Longitudinal clearances serve to test the safety distances. Events, traffic lights, traffic signs, pedestrians, vehicles and their movements in the vicinity of the driver’s vehicle are described within a radius of 250 m (273 yds) around it. Their frequency and the time they appear are calculated and the status of the vehicles’ brake lights and blinkers are indicated. For the introductory trip in the simulator, the overall space used (the combined integral for all deviations from the lane) has proved a good teaching aid. This is calculated from the amount by which the side and central lines are exceeded per time unit in m (yds), by adding them up across the whole simulator trip; it describes the space used outside the lane. Low values for longitudinal, lateral and yaw acceleration, as well as for roll, pitch and drift angles, show how smooth the trip was and are necessary for the evaluation. If the trip is uninterrupted, there should be little steering activity. Other proven performance criteria for driving quality are the Time to Line Crossing (TLC, Käppler, 1993c) and the Time To Collision (TTC, van der Horst, 1990). Both values are based on statistical projections of the driver’s and other vehicles’ paths. While the projection is carried out using vehicle dynamic models, they keep the current driving data at that time constant and are calculated for each dataset recorded, e.g. every 10 ms. TTC is the time which would pass until, under these theoretical conditions, a vehicle would cross the path (also projected) of a moving

98


Table 4.12 Examples of visualization during the debriefing Subject described

Visualized using

Dynamic vehicle forces Shifting centre of gravity Static friction coefficients

Show effects using arrows Show position of tipping edge and tipping Effects of changed coefficients

object or would hit an unmoving object. TTC is calculated for all objects in the vicinity of the driver’s vehicle and is a validated measurement of driving quality. TLC is the time which would pass until, at constant speed and with the steering wheel held tight, the position vector of the driver’s vehicle would cross a line, e.g. the central line or side boundary of the lane. TLC is also a driving quality measurement which has been validated in experiments, and is calculated for the driver’s vehicle for each dataset. It shows the time a driver has, with no further activities, until crossing a boundary. TLC should be as long as possible. Also, engine speeds and brake temperatures should be low, as should the fuel consumption. Experienced drivers can be spotted from their infrequent gear changes and pedal movement, and their frequent use of the engine brake, retarder and cruise control. At the end of the trip, the driver is asked to estimate the difficulty of the tasks, in a questionnaire. All recorded data are transferred to the database. It is a good idea to use other visualization methods for explanatory purposes during the debriefing. This applies to the demonstration of the effects of dynamic driving forces on the vehicle, e.g. when driving around a curve, or the different static friction coefficients with different types of weather or pavement grip, as well as shifts in the center of gravity caused by slipped or poorly positioned loads, see Table 4.12 Examples of visualization during the debriefing. For this demonstration, computer-assisted instruction methods are particularly effective.

4.6.1 Notes on Evaluating the Values Simply recording and visualizing the data and compiling measurements is not, however, enough. An assessment of driving quality is based on the: • Quality of the values and the characteristics to be achieved on the course • Quality of the training programs themselves, i.e. the situations and tasks. Let us begin with the quality of the values. Driving quality is a multi-dimensional mental construct which cannot be described by any individual, known combination of variables. As it is not clear from the start which value actually allows driving quality to be evaluated, and is a criterion for driving quality, more than one value must be taken into account. Due to this multi-dimensionality, with all the many possible values, there is no chance, in practice, of finding combinations of values which can help decide for certain whether a specific value is a valid predictor of driving quality. For this reason, today we act on the supposition that subjective assessments, e.g. by experienced, expert driving instructors, are the best prospect for solving the


99

problem. After all, one characteristic of these assessments is that people are able to perceive many different features, weight them and condense them into a single assessment, e.g. good – medium – bad. As long as they do this formally using a reasonable scale, and a reliable assessment process, this assessment of driving quality can actually be considered a driving quality criteria, see Käppler (1993c; Godthelp & Käppler, 1998; de Vos et al., 1999). The criteria problem has been an issue in driver training and testing for many years. For this reason, for a long time and in many studies, people have been searching for physical values which show high, reliable correspondence with the assessments. If the validity of this correspondence were proved, these values could be used instead of the assessments. For simulators and the type of data they produce, especially, this may be one, or perhaps even the advantage they have over expert assessments. Unfortunately, however, due to the multi-dimensionality of driving quality, a large number of physical values need to be taken into account. Their significance frequently remains a mystery, and people often talk of data graveyards, see above. If more values are now introduced to explore the relationship between assessment, as a criterion, and the value which could replace it, these new values may also correlate with the criterion and with one another, meaning that the search for predictors can remain mired in a tangle of intercorrelations (see Bortz, 1977; Käppler, 1993c; Godthelp & Käppler, 1998; de Vos et al., 1999). The number of correlations which need to be taken into account to correctly interpret the relationship rises so rapidly that with only 20 variables, for example, 190 correlations would need to be interpreted. Analyzing the relationships between these features visually soon exceeds human processing capacity. To help solve this problem, there is a factor analysis process, Principal Component Analysis (PCA). Based on a correlation matrix, it categorizes values according to their correlational relationships with one another into a small number of independent groups and arranges the network of variables in such a way as to explain the intercorrelations between the variables discovered. The PCA generates a synthetic secondary variable, a principal component, which correlates as closely as possible with all the primary variables. To interpret the remaining correlations, further principal components are singled out which are independent of the others. This reduces the relationships further and is carried out until the remaining correlation is insignificant. This results in mutually independent principal components to replace a large number of values which correlate with one another to varying degrees, with no critical loss of information. The main focus here is the question of how many individual, independent factors the network of variables involves, i.e. its dimensionality. The factors are rotated to maximize the explanation of the variances for each principal component. As in regression analysis, PCA finds linear relationships, describing the factor values (the projection of the original variable values on components) and factor loadings (the correlations between the principal components and the primary variables). Their square/plot shows the variance shared by a variable and a principal component. The communality, the sum of the squared loadings for a variable, summed across factors, represents the proportion of variance in a variable explained by the components.

100


Here, in the spirit of promoting the use of simulator training courses, the following procedure is advocated. In the first evaluation loop, all the available measurement values are evaluated. They are then subjected to data reduction, e.g. using factor analysis, as described. Redundant data containing the same information are ignored in the evaluation loop which follows. At the same time, assessments are made, e.g. by the driving instructor, and correlated with the measured values. These loops are re-run until valid values are produced to replace assessments of driving quality for each driving task. One higher target which is difficult to depict is the correlation between these ascertained values and external criteria from real-life driving situations. To do so, values from the real situation are placed in relation to the predictors from the simulation using regression equations. Statistical test procedures are used to discover how likely it is for the correlation ascertained between the two to be coincidental. If this coincidence probability falls short of a preset significance limit, it can be assumed that the correlation has predictive validity for the sample. If the sample of drivers, vehicles and tasks investigated is representative of the overall population, the model correlation which has been discovered can be applied to its subsets. Further notes on the procedure and on statistical validation of the factorization model, e.g. with multivariate variance analyses and F values, as well as examples of its application, can be found in Käppler (1993c). The next part of the evaluation strategy is carried out as follows. Inexperienced test subjects (novices) and experienced drivers (experts) are repeatedly studied with the same conditions and tasks in the simulator. Variables are ascertained and analyzed to determine the secondary variables and their acceptable variation ranges, reflecting the obvious difference between novices and experts in a statistically significant, valid and reliable manner. During repeated evaluation loops, a database of valid secondary variables is produced, describing the driving qualities and operating qualities of ideal drivers (experts) and novices. On one hand, this is used for an initial analysis: Is the test subject a beginner, or experienced, and what level of difficulty should he work at in the simulator? On the other, these values are used to evaluate the course results: Has the driver achieved the training target, or does he have to go through another sequence? Thus, ideal values and more precise ranges of variables are gradually determined, which, over a long evaluation process involving several evaluation loops, are turned into values and are used to differentiate between novices and ideal drivers. If the values are validated, they can be applied to evaluate training course results automatically and for interactive course management. An additional point is that the discussions about driving technique to be carried out in the debriefings are essentially based on an extensive database made up of different interaction parameters. Driving samples gained from experiment provide ideal values for optimal courses of action, which the course is structured around. In future, defining these ideal values and optimal ranges, in order to objectivize the database, will be key tasks for the users of simulator-assisted training concepts. The message has already come across that the situations and tasks play decisive roles in the quality of the course results, and that they need to be subjected to constant evaluation. The training course results can only be valid in reality if the training tasks and driving situations are representative of tasks and events in real-life driving.

4.7 Example Schedule and Simulators

101

4.7 Example Schedule and Simulators This capital contains an organizational timetable for the simulator-assisted training courses which have been described in this chapter and were tested in practice. The following basic prerequisites must be available: • Sufficient number of driving simulators which were designed according to the requirements of training courses described earlier • User interfaces which can be used to put together specific, standardizable scenarios using selectable events. • Ways of collecting and analyzing relevant statistics • Rooms and media for carrying out group lessons. A specific truck driving simulator was designed and produced. The technical specifications were derived from the training course requirements described above and resulted in a dynamic driving simulator with near-original bodywork, a fixed screen and 6 degrees of freedom shown in Figs. 2.24 and 2.25 above. This type improves face validity for the truck driver clientele. It has rather realistic sound simulation and field of view with 180 degrees plus rear mirrors suited to perception of objects behind the driver. The electro-hydraulic motion base offers motions in three dimensions with accelerations limited to 0.5 g. Depending on image content and complexity the specific visual system can process and display up to 3,000 elements at a time. The subsequent box shows an overview of the training program. The organizational timetable is based on the following boundary conditions related to space and equipment: • • • • • • • • •

3 simulators are available at any one time 18 participants at most per day of the course 18 participants sorted into 3 groups (I, II and III), each of 6 people Group start times staggered First group starts at 8 am, the third at 9.30 am Total duration of program is approx. 10 h per group Breaks fitted individually between the group training blocks Each group has the same trainer for one day of the course Each group allocated to a fixed classroom.

The structure of a daily schedule of this kind is shown in Table 4.13. This schedule was used for the Economic Driving program, for Anticipatory Driving Technique and Self-Control, or for a mixture of all three, and is based on strict categorization into time blocks of 45 min. The shaded cells in the schedule denote the simulator training trips themselves. The white cells are the theoretical parts of the course: the group training, the briefing and the debriefing. The briefing, trip and debriefing simulator training elements are in bold. In the first column, the start and finish of each training block is shown. In columns 2, 3 and 4, which follow, the training subject matter for groups I, II and III is described.

102


A brief explanation of the process. In the first block, the participants are greeted in their classroom. The daily agenda is introduced and there is an introduction to how to operate the simulator in Briefing 1. The first practical driving section in the simulator, Introduction & trip 1, is mainly for participants to become familiar with operating the simulated vehicle. The routes are to be selected and set out accordingly. With six participants per group, there are approx. 20 min driving time per participant. In the Group Training blocks which follow, 2.1 and 2.2, unanswered questions are dealt with and initial impressions of the trip are exchanged. Next, in the debriefing, causes of accident and ways to combat them are explained and the topic is discussed. Finally, the drivers are prepared for the next driving task in Briefing 2. In the next exercise, Trip 2, the first training situation takes place. Group Training 3 evaluates the reactions, explains topics which have not been explained or are still unclear and introduces the third training trip. This is followed by a lunch break. The next training situation is in Training Trip 3. In teaching blocks Group Training 4.1 and 4.2, there is a follow-up to training trip 3, the participants’ behavior is documented using the data and the topic of the last training trip is discussed. In the last debriefing, this trip (Trip 4) is then evaluated and analyzed. Between each section, e.g. when moving from the driving simulator to the classroom, enough time and opportunity must, of course, be planned for coffee breaks. In the final “Preview” section, the participants are given handouts with written documents and the analysis data for their trip. This unit, at the end of the day after the simulator trips, is held with all the participants and lasts 40 minutes. The training day lasts about 11 h in total. Overview of the Training Program • • • • • •

Preparing the participants Group training to convey information Briefing to introduce participants to the exercise task Driving practice in the simulator Debriefing and replay of processing and evaluation of training outcome Follow-up activity with the participants.

4.8 Group Training As well as drilling driving in practice, group training courses open up a wide range of opportunities. They act as bookends to the training program as a whole, and need to take into account conditions relating to reliability, as well as drivers’ needs. Older drivers can be considered more motivated to learn than younger ones. As professional drivers have generally not been in regular schooling for a long time, it must be remembered that they are not as able to concentrate on theory, just like drivers selected by their superiors to go on the course. People also learn at different speeds, so the subject matter needs to be provided in bite-size portions. On the other

4.8 Group Training

103

Table 4.13 Schematic schedule Duration 8:00 - 08:45

Group I

Group II

Greeting, introduction, daily agenda Briefing 1 Introduction & trip 1

- 09:30

Greeting, introduction to daily agenda Briefing 1

Debriefing & Replay 1 Introduction & trip 1 - 10:15

Group training 2.1 Group training 2.2 Briefing 2

- 11:00 Training trip 2

Introduction & trip 1 Group training 2.1 Group training 2.2 Briefing 2

Debriefing & Replay 2 Training trip 2 Group training 3 Lunch break Briefing 3


Training trip 3



Training trip 2 Group training 3 Lunch break Briefing 3

Debriefing & Replay 3 Group training 4.1 Group training 4.2 Briefing 4

Debriefing & Replay 1


- 14:00

- 14:45

Greeting, introduction to daily agenda Briefing 1


- 11:45

- 12:30

Group III

Debriefing & Replay 2 Group training 3 Lunch break Briefing 3

Debriefing & Replay 3 Training trip 3 Group training 4.1 Group training 4.2 Briefing 4

- 16:15


Debriefing & Replay 4 Summary

Training trip 4


Preview & closing remarks

Debriefing & Replay 4 Summary

Training trip 4

Preview & closing remarks

Debriefing &Replay 4 Summary

- 17:00

- 17:45

- 18:30 Preview & closing remarks - 19:15

104


Table 4.14 Hazardous goods driver training program: Group training units Topic of training unit

Method

Participants

Duration in min

Introduction to the program of the day • Introduction of trainer and participants • Organization and answering questions

Group introductions Group discussion

All, plus trainer team

15

Brief media-assisted presentation Participant group discussion Brief presentation with Q & A

All

35

Group discussion with group decision

Groups of 6 participants each

70/75

Role play Discussion in small groups Brief presentation


60/40

Brief presentation Brainstorming session using “Metaplan” cards Group work with moderator Group discussion


40

Short group discussion with moderator Participant group discussion Questionnaires

All

40

All, plus trainer team

10

Sensitization for the topic • Current situation in hazardous goods transport • Course aims • Explanation of course unit contents • Relationship between simulator-assisted units and group training Dangers of the job and improving safety • Identifying risks • Possible solutions • How individuals can have an effect Communication training • Improving communication • Asserting your own point of view • Stress reduction Stress and stress management • Focus on participants • Causes of stress • Categories of stress • Practical tips

Integration of training units • Relationship between simulator-assisted and group training course • Incentives for self-reflection • Encouraging transfer from course to real life • Post-analysis Closing remarks

4.8 Group Training

105

hand, drivers taking part in the course voluntarily are sure of their ability to control the vehicle and capable of proving their driving abilities in front of others. The course is peopled with drivers who consider their own skills above average. The general point is more for predictable subjects with practical aspects to be integrated permanently into drivers’ behavioral repertoires. With traffic safety in mind, additions to behavioral patterns which drivers already master are seen as a useful, necessary improvement of their own skills. This applies both to the overall effect of the program and to the trainer’s teaching ability, see Notes On Trainers’ Qualifications. For this reason, the characteristics and expectations of the target group need to be ascertained before the course, using questionnaires, so that at least a minimum is known about the participants’ individual needs and expectations, e.g.: • • • •

Age and social status Schooling and vocational training Previous knowledge, driving practice and experience Truck type driven every day and daily routes.

As part of the simulator course for professional drivers described above, group training lessons take place on different topics. Table 4.14 below describes the group training units for the Hazardous Goods Driver Program; the first column shows the Topic of the training unit; the second, Methods; the third the Participants involved, and the fourth, how long it takes.

4.8.1 Introduction to the Program of the Day The introductory session with the trainer team and all the participants serves as a group discussion to introduce the trainers and participants, to present the organization, responsibilities and schedule, as well as to answer any questions.

4.8.2 Sensitization for the Topic The aim of this training unit is to: • • • •

Lead up to the problem Encourage motivation for the course Get to know each other Create a positive atmosphere.

In this training unit, with all participants, at the start of the day, the topics listed in the box are dealt with in two short presentations and a group discussion.

106


Sensitization For The Topic Brief Presentation Situation in Hazardous Goods Transport, 5 min • • • •

Working conditions Competition Legal guidelines Public criticism

Participant Group Discussion 15 min • • • • •

Expectations about the Course Today,

Set down results on the flipchart Aims of course Explanation of group training course units Introduction to simulator-assisted training Answering participants’ questions

4.8.3 Dangers of the Job and Improving Safety In the second training unit, the topic is existing dangers and traffic safety improvements, as well as how to handle hazardous materials safely. The method used for this is participant-centered and aims to: • Identify dangerous situations at work • Develop possible solutions with the participants, and • Recognize ways individuals can have an effect and change things. The method is group discussion with a group decision, as used in safety work (Brehmer et al., 1991; Misumi, 1978, 1982) in groups of six in three phases, see box below. The trainer: • Informs the participants at the start about what the aim is and how the schedule will run, and ensures events occur as scheduled • Encourages all group participants to take part in discussion and makes sure they all have a chance to speak • Ensures, during the first phase, that the points noted down are succinct, and are later understood in the way the person in question intended them • Leads the topic of the discussion and relates to the participants’ experience at work and circumstances • Does not express his own opinions. In this process, safety-critical situations are identified by the participants themselves; that is, individual aims are taken into account, enabling long-term behavioral changes in concrete situations.

4.8 Group Training

107

Dangers of the Job and Improving Safety Phase 1 Topic Discussion and identifying problems and stressful aspects of everyday working life when transporting hazardous goods, which can have an effect on safety. Implementation In the introduction, we already touched on general risks, problems and stressful situations during the transport of hazardous goods. You are the on-the-spot experts: as hazardous goods drivers, you deal with all kinds of different situations every day and usually have to react and make decisions very quickly. On this training course we have the opportunity to take a look at specific problem areas from a certain distance, so I would like you to list and discuss, together, general problems in your work which can affect safety. You have 25 min to do so. Please note all the subjects mentioned on a card, so that none are lost. Phase 2 Topic The problems identified in Phase 1 are added to by the other groups and put into two categories: • Problems which can only be solved by the company management alone or with its cooperation • Problems which can be solved by the group members themselves taking action. The moderator takes the first list and indicates that the points will be passed on to those in charge. This can take place at seminars for decision-makers. For a positive overall outcome, the participants must receive feedback. Implementation Before the break, you made a list of problems you face every day and wrote them on a card. I suggest taking a closer look at those cards and would like you to sort them into two categories. There should be one pile for cards with problems that can only be solved in cooperation with the company management, only by the company management or by other departments. The second pile should be for the cards listing problems you yourself can solve, as an employee. Phase 3 Topic Discussion, as concrete as possible, of problem areas in Phase 2 which can be solved by the employees themselves. Phase 3 identifies the drivers’ own behavior needed to solve the problem. It is not enough for the discussion to end with the general resolution “I will behave more safely in future.” The aim is for it to lead on to concrete examples which show what safe behavior consists of: “I will get up earlier in the morning, so that I do not have to drive so fast to arrive on time” or “I will plan my work better at the start of the day.” Finally, the group members are asked to decide for themselves

108


what they will do to work more safely in future. They write their decision on a piece of paper, which they put in their wallet or purse without showing to anyone. Implementation In the last step, we sorted your everyday problems which can affect safety into two different categories. I will keep the cards with problems that can only be solved in cooperation with the company management, only by the company management or by other departments. We will collect these problems and let your company know about them. Now let us turn to the problems you sorted into the category “We can change something about these problems ourselves.” I would like you to discuss these problems in the group and think together about what can be done to solve them. Please try to find extremely precise practical ways of solving them. You have half an hour to do so. After you have discussed what precisely you can do to behave more safely, I would like every one of you to decide what you yourself want to do to improve safety. Please note your decision on a piece of paper and put it in your wallet without showing it to anyone else.

4.8.4 Communication Communication is important: forms of communication may play a decisive role in training, too. There are two different ways of depicting objects: using analogy (drawings, gestures) and signs (names), see box. The connection between the name and the object is coincidental; people simply agree on the meaning. The two forms of communication complement one another, as the relationship aspect must be supplemented by words to make it open to linguistic communication. Communication using names has a meaning which does not convey the relationship. On the other hand, the meaning of analogy is not sufficiently clear. Every act of communication takes place on two levels: the content and relationship level, whereby the latter classifies the former. Whenever one person says anything to another, he or she is simultaneously defining his relationship to the other. There is a special vocabulary for both levels of every communication: one is mathematical and concrete, the other is graphic and emotional. Communication is meant to transfer information: subject matter and facts; these, at the same time, contain hints about how the information is to be understood. There is a relationship between the content and the relationship aspect: the latter shows how the data are to be interpreted.

4.8 Group Training

Digital depicting Processes figures facts data logical conclusions rules, laws theories specialist terms abbreviations symbols Analyzes Proceeds step by step Measures time Finds general rules Determines objectively

109

Analog depicting Sees examples images concepts visual material analogies sketches size comparison shapes sounds Is without words Recognizes intuitively Processes holistically, in a nonlinear way Uses body language Has no sense of time Creates new conceptual connections Is creative, imaginative Pictures mentally

A message on the content level does not, however, necessarily have to correspond to the message on the relationship level. When these messages contradict one another, this is a paradox communication. In the following, an approach is presented which enables communication processes to be analyzed in a relatively simple way. To make yourself understood and to carry out plans of action, it is necessary to respond to others and put yourself in their shoes. Meaningful interaction needs people to be capable of role-taking. Role-taking makes people take the views of all the other parties into account. Leadership positions, for example, not only require complex or organized role-taking, but also forms of role-taking which are more neutral, objective and based upon rules and justice.

Nature gave us one tongue and two ears so we could hear twice as much as we speak – Epictetus Company discussions are very often characterized by a lack of willingness to role-take. The main reasons for this are conflicting aims and values: For example, the values of the management and the employee organization differ, or those of superiors and subordinates. The role standpoints brought in by the parties are also significant. One relatively simple means of role analysis is transactional analysis. This model is based on the three ego states:

110


• Parent ego state, characterized by rules, bans, standards and moral concepts • Adult ego state, characterized by reason, objectivity and a rational, detached manner • Child ego state, dominated on one hand by spontaneous, creative, inventive feelings, and on the other by conformist, unprotesting behavior. In the course of socialization, individual behavior patterns develop which cause different ego states to dominate. In conversation, these ego states come up against one another and guide the interaction. In transactional analysis, a distinction is made between complementary transactions, which are not a problem for interaction (where a stimulus excites an expected reaction by the other, in the ego state addressed) and crossed transactions, which are problematic for interaction (where a different ego state is activated to that addressed). A crossed transaction occurs, for example, if a speaker reacts to the other’s comment in an unobjective, reproachful, critical, derogatory manner. In a crossed transaction, the interacting people adopt ego states which bring out emotions. With these transactions, either the subject is changed or the interaction is broken off. The following rules have stood the test of time: • Do not try to force your information across on the level of content if the level of relationship is damaged • Instead, speak on the level of relationship; here, your body language (expressions, gestures, tone of voice) is especially important (calming, balancing). As long as the other person is lost in a psychological fog, he has little capacity to take in the current information and mainly perceives signals on the relationship level • As soon as this fog has cleared, communication can continue on the content level. Example of a Crossed Transaction between Persons A & B A: Just pull yourself together! B: How can I help it if you explain your stupid exercise so badly? In this example, each person is appealing to the other’s child ego state from a parent ego state position – a crossed transaction situation which leads to trouble. The important point is that the crossed transaction situation is recognized when it occurs, and each person gives up his or her parent ego state in favor of an adult attitude. This could be as follows: B: I’m sorry, but I don’t feel I can cope with your task. Could I continue with a simpler one? Even if Person A plays on for another round (It’s simple; a child could do it!), in the long term it is hard to avoid a complementary transaction on the adult-to-adult level, the adult ego state.

Good listening is at the heart of good communication. We never listen in a disinterested manner. Our experience becomes mixed up with what we hear and changes

4.8 Group Training

111

it. This change is often not conscious. Our expectations affect how we understand what we hear, and we become involved in what we hear, with our emotions and feelings. Words can call up distinct feelings in us, which, in turn, lead us to interpret things accordingly. We often do not hear everything. Thoughts can be distracted. When you speak to others, look out for the following points: • Use factual statements • Avoid making judgments, attacking or putting people down • Your interlocutor’s processing capacity is limited: do not introduce too much information at once • Use your best argument at the start or the finish, not in the middle • Respond to others (Yes, I agree with you fully on that point, and would like to add that . . . ) • Ask your partner what he or she understood.

Rules of Good Listening Do not speak You cannot listen if you are speaking Relax the other party Show the other person that he or she can speak freely. Create a tolerant atmosphere Show that you want to listen Show interest. Do not read your mail during the conversation. Do not doodle, pile up papers or leaf through documents. Listen to understand, not to oppose. Avoid distractions Would it be quieter with the door shut? Adapt to your partner Try to put yourself in his or her position, so that you can understand the other’s point of view. Patience Take your time, do not interrupt, do not hover Stay in control If you are annoyed, you will understand the other person’s words wrongly. Do not be thrown off-balance by reproaches This forces the other person to react. Do not argue. Even if you win, you have still lost. Ask questions This encourages your partner and demonstrates your interest. It can intensify the conversation. Do not speak This is the first and the final rule, and all others depend on it. You cannot listen when you are speaking. In communication training, participants practice positive management of communicational conflict situations. The central theme is dealing with customers, as a concrete real-life example, see Communication box. The training unit is carried out in groups of 6 and lasts 40 min. The aim is all about: • Recognizing and improving your behavioral pattern • Asserting your own point of view • Reducing communication-related stress.

112


In the communications course, role plays are used to demonstrate a negative model, as well as work in small groups and brief presentations. Communication 1. Role play, 5 min Situation Roles 2. Group work, 5 min

3. Discussion of the results, 5 min 4. Work in small groups, 10 min 5. Presentation of the group results, 10 min 6. Conclusion and Q & A, 5 min

Conflict with customer at gas station The trainer is an aggressive customer The participant is a stressed driver Put a table on the flipchart What precisely happened? What made the conflict escalate? How was further escalation avoided? Aspects of verbal and non-verbal behavior Developing a positive behavioral model Two groups are formed. Each group works on a situation in the customer-driver role. Each group plays out the situation it worked on

4.8.5 Stress and Stress Management The topics of this training unit are stress and strategies to deal with stress. It is carried out in groups of 6 and aims to: • • • •

Let participants know that their problems are taken seriously Show participants that their own situation is not an exception Identify sources of stress Categorize stressful situations according to how much influence the participant has on the stressful situation, and the source of stress (participant or others) • Provide practical tips on how to manage stress in practice The methods are: • Introductory brief presentation with basic information about stress • Brainstorming session using “Metaplan” cards, to identify stresses at participants’ work and elsewhere

4.8 Group Training

113

• Group working during which the identified problems are sorted into categories • Group discussion about ways of managing stress Stress and Stress Management Brief presentation, 5 min • Types of stress • How can we tell when stress is arising? • Causes of stress Brainstorming session using “Metaplan” cards, 10 min • What stressful situations do you experience in everyday life? Group work, 15 min Sorting the cards into categories: The problem can . . . • • • •

be influenced by me not be influenced by me be influenced by others/from outside not be influenced by others/from outside

Discussion of the results and imparting information, 10 min • Means of stress management • Link

4.8.6 Integration of Training Units Here, special emphasis is placed on exchanging experiences gained during the day of training, and integrating the individual training units. The aim of this training unit is to: • Ensure participants see how the content of the training units fits together • Reflect on the content of the training unit • Encourage transfer of this content to everyday situations. Short, moderated discussions, group discussions and questionnaires are the methods used, see box.

114


Training Unit on Integration of Training Units Group discussion, 20 min • To what extent can the knowledge and skills gained during the course be transferred to everyday working life? • In which areas of everyday life can concrete changes be made/achieved? • Do you have the impression that you are better prepared to react to critical situations thanks to the simulator trips? If so, which situations? Group discussion Achieving aims, 10 min • To what extent were your expectations fulfilled by the course? Questionnaires, 10 min • • • •

Assessment of the day’s training Assessment of the trainer Impression of the driving simulator Impression of the driving tasks

4.9 Questionnaires At the end of the day, questionnaires (Käppler, 1993c) are passed round. The information is also the basis for the assessment of the course and improvements to the training program.

4.9.1 Assessment of the Day’s Training The first questionnaire targets assessments of the day’s training with questions about: • • • •

Organizational procedure Group training Simulator trips Recommendations.

In the questionnaire, statements are made about how the training just finished went, and your impressions. Please evaluate the day’s training according to the following statements and fill out how accurate the statements are using the scale. Please decide if for you the following statements are not true, or very true. Cross the place on the scale that fits – even between the markings.

4.9 Questionnaires

115

My expectations about the course today were not 0

a little somewhat

1

2

3

partly

rather

mostly

very

4

5

6

7

8

fulfilled. In respect to my work, I found the course not 0

a little somewhat

1

2

3

partly

rather

mostly

very

4

5

6

7

8

useful. I found the organization not 0

a little somewhat

1

2

3

partly

rather

mostly

very

4

5

6

7

8

good. I found the premises not 0

a little somewhat

1

2

3

partly

rather

mostly

very

4

5

6

7

8

appropriate. I found the food not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

good to eat. I found the group training not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

useful. I found the simulator training not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

useful. I will not 0

1

a little somewhat 2

3

partly 4

rather

mostly

very

5

6

7

8

recommend the course to colleagues.

4.9.2 Assessment of the Trainer The second questionnaire is to assess the trainers and their work in the group: In the following questionnaire, statements are made about the trainers and your impressions of them. Please evaluate the following statements and complete them using the scale. Please decide if for you the statements are not true, or very true. Cross the place on the scale that fits, even between the markings.

116


I found the trainers not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

easy to get on with. I found the way they interacted with the participants not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

appropriate. I found the way they motivated us to join in not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

appropriate. I found their method of presentation and speaking not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

appropriate. I found their use of working materials not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

appropriate.

4.9.3 Semantic Differentials for Driving Simulator and Driving Tasks Semantic differentials are used to assess both the simulator and the driving task. These semantic differentials, or bi-polar scales, used in advertising psychology and acceptance research, measure people’s personal attitude toward objects. Here, the emotional, associative meaning of a term or object, rather than its definitional meaning, is measured and tied in with the respondent’s different, individual experiences. The semantic differential offers the advantages of being simple to use, widely accepted and reliable. In the case in hand, the simulator and the driving tasks should be assessed separately using semantic differentials. These consist of 15 pairs of adjectives in the following boxes, aimed at assessing: • The simulator, in terms of technological and subjective aspects, and • The driving tasks, in terms of realism, practical relevance, interest and level of difficulty. The data compiled are used along with the measurements from the simulator to evaluate and improve the course. Before the semantic differential is filled out, instructions are passed out, see below. The participants are informed that: • There are no right and wrong answers • They need to make sure to check all pairs of answers • Every word pair can only be checked once

4.9 Questionnaires

117

• Every word pair is to be checked independently of the assessment for the other pairs • They are to work through the word pairs quickly and not think for long or mull over the answers, as the point is to record their first impressions.

Instructions for Semantic Differential on Driving Simulator and Driving Tasks

As we are interested in your personal opinion about the simulator and driving tasks, and it can provide us with important ideas, we would like you to fill out the following questionnaire. It is made up of pairs of opposite words. We would like you to assess first the simulator and then the driving tasks, separately, using the listed words. Record the level of your opinion in each case. To do so, boxes are drawn between the two words. Let us take the example of the word pair interesting–uninteresting. If you believe that the driving tasks are very interesting, make a check next to interesting. interesting

uninteresting

If you believe that the driving tasks are very uninteresting, make a check next to that. interesting

uninteresting

If you believe that the driving tasks are interesting or uninteresting, make a check in the second box away. interesting

uninteresting

interesting

uninteresting

If you believe that the word interesting applies better than the word uninteresting, make a check in the third box away from interesting. interesting

uninteresting

interesting

uninteresting

Use the middle box if the driving tasks are neither interesting nor uninteresting.

118


Semantic Differential for Driving Simulator Think back to the driving simulator. The word that goes with it is: unpleasant safe stays on track revolting convenient unrealistic cheap uncomfortable useless fun tough

very rather somewhat medium somewhat rather very pleasant unsafe swings out attractive inconvenient realistic expensive comfortable useful boring delicate.

Semantic Differential for Driving Tasks Think back to the trips in the driving simulator. The word that goes with those is: individual unusual every-day high-risk unsuitable unrealistic varied easy too challenging

very rather somewhat medium somewhat rather very general usual unique low-risk suitable realistic monotonous difficult not challenging.

4.9.4 Preparation Before they first take part in the course, the participants are introduced to the forthcoming training course. This is done by means of written information and supporting material sent by mail: • About the training center, the aims and procedure of the daily training schedule, and how to get there • Questionnaires on participants’ private and professional situation to a certain extent, on their attitude toward road traffic safety, and on their driving style. The participants are assured that this information is used solely to optimize the way the course is run and the evaluation that follows, that it will be treated as confidential and not passed on to third parties.

4.9 Questionnaires

119

4.9.5 Private and Professional Situation The information on the participants’ private situation enables the content of the group training units to be adapted to the participants’ situation in life, and helps to sort them into groups. The questions are related to: • Personal information: name, age and marital status • Schooling, school qualifications and vocational training • Driver’s license and other additional permits. The advance information on participants’ work and professional situation is also used to link the course with their everyday experiences and add to the feedback on the trips using practical examples. The information is related to: • • • • • • • • • •

Range of company activities Number of employees in the company Length of time participant has worked for current company Working hours and shift rhythm Daily time behind the wheel Participants’ concrete activities at work Fixed or changeable routes Route characteristics Vehicle equipment (EPS, etc.) Route driven by participant to his place of work.

4.9.6 Questionnaire on Attitude toward Road Traffic Safety and Driving Style There are different attitudes and opinions on professional drivers and traffic. These are of interest before the course, both for its implementation itself (target-groupspecific approach) and to test for changes in attitude after the course. Using the questionnaires Attitude Toward Road Traffic Safety and Driving Style, participant profiles are compiled and information gained about changes in attitude and behavior during the course. The results before and after the course are compared. In the questionnaire Attitude Toward Road Traffic Safety, a list is presented with statements about driving. The statements are taken from the 98-item questionnaire to measure attitude toward road traffic safety (Holte, 1994). It is available in several versions of differing length which have been subjected to tests for reliability and validity. In the case in hand, a 12-item version was selected, developed from a 15-item version by Holte (1994). The box shows the questionnaire Attitude Toward Road Traffic Safety. The respondents agree to each statement to varying degrees by

120


indicating a number on the 8-point scale below, used earlier: Once again, values can be selected between the markings. not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

The questionnaire on driving style is based on the Driver Behavior Questionnaire (Parker et al., 1995) and consists of 15 statements, sorted into the error types of lapse, error and violation. In this version, see box, these three categories remained

Questionnaire on Attitude Toward Road Traffic Safety 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

Driving is fun when you can really put your foot down. For me, a wet street is no reason to drive a lot slower. On secondary roads, I am often held back by drivers going too slow. Driving comes completely automatically to me. I often drive faster than the permitted speed now and then. It’s a good feeling to rev the engine up when accelerating. It is possible to drive fast and carefully at once. If I want to be on time, I do sometimes drive faster than the permitted speed. Even when traffic is heavy I try to reach my destination speedily. I like a challenging high-speed drive. No-one should be allowed to drive faster than 120 km/h (75 mph) on the freeway. I enjoy getting into races with other drivers.

equally distributed. Using the 8-point scale already employed before, the respondents indicate how often in the last twelve months they have demonstrated each type of behavior, i.e. “not often” to “very often.”

4.9 Questionnaires

121

Questionnaire on Driving Style 1. I accidentally approach a traffic light in third gear. 2. I drive close behind the vehicle in front to show the driver he needs to drive faster or make way. 3. I cannot remember the route I have driven on very well. 4. I cross an intersection although the light has already changed to red. 5. I forget to look out for pedestrians crossing the road when turning from a main road into a side road. 6. If I am annoyed by another road user’s behavior, I tailgate him to show what I think. 7. In the late evening or early morning, I do not keep to the speed limit. 8. When turning right I almost drive into a cyclist going along directly on my right. 9. I have an aversion to certain groups of road users and show this whenever I can. 10. If I overtake on the freeway, I underestimate the speed of an approaching vehicle. 11. When reversing, I fail to see something standing in the way and drive into it. 12. I want to drive to place A and realize after a while that I am driving to B as it is a route I drive along often. 13. I get in the wrong lane at an intersection. 14. When setting off, changing lanes, etc. I forget to look in the rear mirror. 15. I brake suddenly on a slippery road surface or make my vehicle skid by steering wrongly.

4.9.7 Follow-up and Training Needs After they take part in the course, continued contact is kept with the participants by mail, to gain information about how long the results achieved last. The questionnaire Attitude Toward Road Traffic Safety, the questionnaire Driving Style and the following scales on training requirements are sent six to eight weeks after participants take part in the course. A few weeks have now gone by since the training course. We would like you to evaluate the course from your point of view. In the following questionnaire, statements are made about how the training went and your impressions. Please check your opinion according to the following statements and complete the statements using the scale. Please decide for you if the following statements are not true, or very true. Cross the place on the scale that fits – even between the markings.

122


In retrospect, for my daily work, the course was not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

useful. Since the course, I see not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

many things I can change. I have not 0

1

a little somewhat 2

3

partly 4

rather

mostly

very

5

6

7

8

recommended the course to colleagues. I would not 0

1

a little somewhat 2

3

partly

rather

mostly

very

4

5

6

7

8

like to take part in another course.

4.10 Notes on Trainers’ Qualifications, Briefings and Replays Previous training research has put the Requirements for a Successful Training Course together (see box below, according to Birkenbihl, 1995). The aim of briefings, e.g. is to clear up any uncertainties about the subject or the organization in advance. Even in the days of open cockpit biplanes, it was absolutely essential to agree upon what to do. While the pilot flew from the rear, the passenger navigated. As the two were hardly able to communicate during the flight, they had to agree in advance on where they were going to fly, on which route or which alternative routes. Briefings have been mentioned throughout this text. The concept of the briefing was honed in the context of military deployment. To destroy specific large areas with a fleet of bombers, those involved needed extensive briefing. At the same time, it was discovered that a debriefing was also necessary to record information gathered about defense positions or hits. Today, commercial airline companies carry out standardized briefings. For example, the weather status, flight time and number of passengers are reported. In the separate working groups, in the cockpit, in the cabin, task allocation is discussed and the style of the approach flight is set down, among other things. Details are given in Käppler 1996c).

4.10 Notes on Trainers’ Qualifications, Briefings and Replays

123

Requirements for a Successful Training Course • Teaching can lead to success if new behavior patterns are acquired and practiced by means of active learning • New behavior patterns are repeatedly practiced. Each repetition must take place in a situation which appears different • Each training unit must be worked through in a manner suited to its content. The thrust of the teaching method must fit the type of material: just one method for all subject areas is wrong • Course participants want to be challenged. The greater the challenge, the greater the success. The participants should be taken to the limits of their ability • The aims and organization of the course must be clear. Start with the known and add to it using examples and comparisons from various walks of life • The material should be provided in context, with a wider overall plan. The wider contexts and logical links between sub-areas must be clear • Fear is detrimental to learning, demotivates and creates mental blocks. The trainer must create an atmosphere without fear and give those who are excessively fearful positive attention • Training must be light-hearted, not deadly serious. If you can arouse curiosity, make the course varied and spice it up with humor, you will motivate participants to learn and change their ways.

This holds true for trainings as well, the trainee needs to be given thorough instructions, supported during the exercise and then informed afterward about how he or she dealt with the tasks. They have to know what they are supposed to learn, have an idea of what demands and difficulties will arise and be able to judge what they are doing right and what not.

124


Prebriefing clarifies the aims, boundary conditions and program of the exercise to come, and serves to explain and demonstrate the procedures and detailed planning of the course depending on each participant’s skills and knowledge answering the following questions: • • • • • •

What is the situation like which you are about to deal with? What season, time of day, truck and load will there be? How long will the tasks take? What will be required, when? What is the main thing to watch out for? What results are expected? We will set off when . . . and then . . . It will be important for you to . . . Make sure not to . . . if possible. If you are not sure, you can . . . If necessary, we can . . . • Procedure X will be introduced by . . . You then . . . in this way. Briefing determines whether a task fits. Deficiencies are dealt with before they disturb the overall course of events. The trainer pays special attention to irregularities, points them out, learns what he needs to watch out for during the exercise and uses this knowledge for replay sequences. This consists in: • • • • • •

Briefly answering questions Briefly encouraging a participant Brief criticism of erroneous or dangerous actions Slipping in critical situations Giving further instructions Stopping the course in case of serious mistakes.

Debriefing works on the results and consists in • • • •

A brief discussion of what happened Comments on the results for reaction times, staying in lane, etc Targeted criticism of content for certain points Underlining the importance of the exercise in practice; pointing out how the simulation may differ from practice • Summarizing the results and previewing the next exercise

The question is: What does the trainee need to find out during the exercise, by trial and error, and what can the trainee be told before, during or after the exercise? The answers to this question depend on skills, motivation and the tasks set. The main task for the trainer is to judge these things. Other requirements for evaluation are, of course, the documentation and assessment of possible actions. A driver’s gear


125

changes in a certain situation, or how well he stays in lane, have to be recorded in order for them to be checked and evaluated. Replays can be used, as well as spoken discussion after the event. The ability to replay stored training trips is one major advantage of simulators and is used for objective, targeted feedback. It shows clearly what went well and where there are faults. Replay sequences help in evaluating participants’ current level of training and planning further exercise steps. The aim is objective proof by demonstrating remaining shortcomings and showing which challenges have already been solved well. Beforehand, good behavior patterns are selected as demonstration material and used to show what the driver needs to pay attention to. In the replay, the way each driver dealt with the task is observed, analyzed and discussed. The procedure is as follows: • Identification and selection of typical right and wrong behavior, demonstrating the differences between good and not good • Demonstration of good behavior • Replay, discussion and analysis of how individual drivers dealt with the task; pinning down weak points. Among other things, training experience and research show that one major factor for the success of the course is the trainer him- or herself (Winkler, 1988). Trainers must be able to manage exercises so well that they make no mistakes which could throw doubt upon their ability. They must be able to imagine processes, to tell the participants the point in the exercise where they made a mistake, or how well they acted. They must be capable of vivid verbalization. Trainers need to be able to react to the needs of a group of adults and support them all day long in an educationally effective manner. The teaching matter must be taught in a graphic way; the participants’ different levels of achievement must be taken into account and gaps bridged. The trainer’s ability decides how well the course goes. His or her behavior and personality direct the group process. It is up to him to encourage open discussion, motivate participants to work through the issue and increase their awareness of problems; the trainer can induce changes in behavior. More details may be found in Beelitz, Käppler et al. (1995), e.g. Thus, the trainer must have professional knowledge and good teaching, social and psychological skills. Professional knowledge is principally related to traffic psychology and what goes on in traffic, to the practice of traffic psychology diagnostics and to knowledge of traffic risks and dangers. Educational skills include knowledge of how adults learn, and experience in work with groups of adults. Social and psychological skills include experience in communication and dealing with group dynamics, and qualification in leading discussions. These enable the trainer to respond to the individual participants’ various situations and experience and integrate this into the group process.

126


Checklist for the Trainers • Start your presentation in a personal, factual manner. Make positive contact with the participants. • Aims of presentation. At the start, say what the aims of your presentation are. • Organization. Start off by giving the audience an overview, introducing the organization of the presentation • Visualizations. Support your main statements using transparencies, slides or film • Stimulation. Make your presentation come alive with specific examples and comparisons • Speak plainly. Use common words and short sentences. • Don’t tell – ask. Have you thought up questions for the discussion with the participants? Make the participants active by ASKING instead of TELLING everything. • Eye contact. Look at the participants. • Positive working atmosphere. Thank people for questions and objections. Enter into questions or bounce them back to the asker • Conclusion. Summarize the key points of your presentation once again as theses, and conclude with a personal comment

The trainer must come across as a competent person who can convey knowledge and has a motivating, activating effect on adults. On the other hand, in certain phases of the course, he needs to guide the participants by trying to enable and help the group to form opinions and intentions, without influencing them materially. Thus, the trainer must be an expert in communication and group processes, who structures and organizes discussions: he is not the group leader or teacher. Established findings from the fields of traffic psychology, driving instruction and adult education show that detailed technical knowledge and advanced driving skills tend to play a more minor role. For details see the Checklist for the Trainers in the box below.


127

Two drivers Mr. K., asked about the approach of two theater directors, compared them as follows: “I know a driver who has the traffic regulations at his fingertips, obeys them, and is able to use them to his own benefit. He is skilful at racing forward and then maintaining a normal speed again, going easy on the engine, and thus he makes his way carefully and boldly between the other vehicles. Another driver I know proceeds differently. Even more than in his own route he is interested in the traffic as a whole and he regards himself as a mere particle of the latter. He does not take advantage of his rights and does not make himself especially conspicuous. In spirit he is driving with the car in front of him and the car behind him, with constant pleasure in the progress of every vehicle and of the pedestrians as well.” – Bert Brecht (1935)

Chapter 5

Concluding Remarks

Different versions of the training program for professional drivers described here have been tested in practice. The insights gained in this process have been used to optimize the program and procedure and were taken into account in this book. The training modules Economical Driving and Anticipatory Driving, as described above, were carried out over two at two training centers in Germany. The Economic Driving course, in particular, showed signs of success and a reduction in fuel consumption. The question currently being asked is that of how reliably these effects will last and how often the courses will have to be repeated, at what intervals. The idea of the Self-Control continued education program lies in the fact that fixation itself is understood, as well as the problems resulting from it. The program, however, was never be tested in practice. Several points have become clear. First, that the effects of group trainings have been under- and those of simulator trainings overestimated. These were even understood by the drivers as teasing supplements to the group trainings. Therefore it is recommended that further development and test be aimed at the effectiveness of group and simulator training combinations. Consequently, it has also become clear that demands upon the trainers, in general driving instructors, cannot be underestimated, and that the qualification needs to be complemented by Train The Trainer seminars and continuous quality checks. One significant advantage of simulation is that irrefutable assessments of driving quality and the success of the course can be attained using objective data. It has become clear, however, that it is not actually known, in the fields either of driving instruction or of simulation, exactly which variables driving instructors use as a basis for subjectively assessing good driving quality. In driving schools, the driving instructors’ skill actually conceals the fact that here, there is a lack both of validated performance criteria and of corresponding permitted minimum or maximum values. As the system more or less requires driving instructors and their skills to be absent in simulators, one of the main requirements for the successful use of simulators is for performance criteria and correspondingly well-accepted ranges of criteria to be created by means of research and development work to this effect.


129

130

5 Concluding Remarks

Some further critical comments. Learning targets related to social behavior involve the road users’ behavior within the group. In simulator centers, several simulator units may be joined together in networks so that several drivers can actually interact in one traffic network. This enables them to practice anticipating critical traffic situations, overtaking and how to behave at intersections. As yet, the tantalizing question of what learners gain when taught with learners, and how, remains unanswered. The social interaction learning targets have also been achieved virtually using homunculi. Every virtual road user acts independently and interactively, within certain limits. The structures and functions of machine learning allow continuous, changing processes of interaction between the virtual road users and the subjects in question. Every model virtual driver constantly makes new decisions about what to do next, taking into account all the other drivers and the specific surrounding conditions. However, considerable development work still has to be carried out to create multiple behavior types with a wide range of different unpredictable decisions and actions in further traffic situations. This modeling of human behavior is not yet realistic enough for the behavior, interactivity, range of tasks or variations to be experienced as sufficient or valid. After sequences of repetition, models are easy to double-guess, meaning that the homunculi’s reactions can be predicted. It is clear that for different tasks from different fields, very different simulators are the ideal means of training. Traffic structure, acceptance, emission reduction and availability of vehicles must be taken into account when considering the technological and operational requirements for driving simulators. The teaching matter, media and technological and operational requirements must be adapted to fit together. For this reason it is recommended to use the 4 class technological taxonomy of driving simulators in Fig. 2.12 and in Table 2.4 at the, known as a simulator family. These may consist of: 1. Full-task simulator for drilling critical dynamic driving situations such as evasive maneuvers, i.e. number 6 in Fig. 2.23 2. Normal driving simulator for normal, less dynamic, low-event trips such as on freeways or secondary roads, i.e. number 5 in Fig. 2.21. Such simulator type has been used for the Economic and Anticipatory Training Programs in this book 3. Procedural simulators with highly simplified yet still complete representations of driving tasks such as learning to navigate through a city, i.e. number 3 in Fig. 2.17 4. Part task trainers to teach isolated subtasks such as hill starts, i.e. preferably number 2 in Fig. 2.15 or with further restrictions number 1 in Fig. 2.13 This leads to a complete smart driver training media family if this taxonomy is rounded up with computer-assisted instruction, training and test (CAI and CAT) to teach and practice using the simulator, the tasks and the rules, as well as handling procedures and to run the tests. It is consequent that Germany allows for the theoretical driving tests using CAT. Using all these media together, a self-contained teaching system can be achieved if it is systematically rounded off by vehicles on the road. This procedure is less expensive than carrying out all tasks, right down to the simplest ones, in one


131

single simulator, which would necessarily have to be very high-performing and would sometimes have to be used below its full potential, at great expense. However, this approach requires greater effort in designing the course, as, for every single task, requirements for technological training media and organizational matters would have to be taken into account or set down, intermeshed with the requirements and aims of the course. One incidental aspect of simulator-assisted learning is a detachment from the real, concrete situation in question. Simulators distort teaching or training to a certain extent by detaching the learner from the actual, real situation and its risks and dangers. When dangers and risks are not present, however, people do not experience the consequences of their own actions, or only do so insufficiently. In combination with excessive practicing of sensorimotor skills, this can lead to an exaggerated sense of their own skills, an uncritical feeling of safety and uncertain motivation. This is undesirable. Own experience with the Training Programs and Janssen (1988), for example, assert that more attention needed to be paid to subjects’ motivation in simulators. As well as the a priori motivation, which is often unknown, he believed that the current level of motivation is also significant. He therefore suggested mechanisms to reward or punish people for their current behavior using (mainly monetary) gains or losses. The aim, he said, is to teach the desired behavior and provide realistic motivation. For example, in a long-range study for a distributor, Tschernitschek (1978) showed that when drivers were rewarded for not having accidents, the accident rate sank by 75 percent over a period of 26 years. There are a range of reports of similar measures from industrial companies. Hale & Glendon (1987) did, however, show that the effects were less than conclusive, and claims were often not backed up by studies. In view of the impressive results, this seems unfortunate. However, not all kinds of motivation can be achieved using money, and it remains uncertain how real the level of motivation simulated in this way actually is since it can still be affected by time pressure, for example. The Sandwich Design, for example, interweaves the study or teaching plans so that test subjects in a group have to complete certain sequences together during one morning, to motivate one another. Another way is to motivate people by means of instructions, such as Imagine you are late for a meeting, or are a very busy businessman. The idea behind these instructions is to standardize the motivational factors. Improved traffic safety generally requires people to be conscious of how dangerous road traffic is. For this reason, all attempts to improve safety and reliability should be accompanied by campaigns to propagate the desired behavior, aiming at long-term changes in attitude. It is still necessary for the way in which attitude and behavior is meant to change to be clear, consistent and accepted. By using modern technologies intelligently, an increasing range of task designs is possible, with flexible scope for decision and action. This is described as learner autonomy and democratizing the learning process: Opportunities to do this have not yet been exploited. Driving simulation has always been, and still is, markedly oriented toward technology. This orientation will change and grow conceptually. The task contents, the way learning is organized, communication about learning

132


and the hierarchies between the teachers and students are being adapted to safety and traffic requirements. They set the requirements for simulation and simulators, rather than vice versa, as is often still the case today. There is therefore agreement on the point that alternative learning and teaching concepts need to be developed which make optimal use of the advantages of collective action, while taking appropriate steps to avoid possible disadvantages, such as practicing in a group of novices. Corresponding multi-level concepts in driving instruction which could be carried out using simulator families were already suggested by Brown et al. (1987) and Duncan et al. (1991). For example, basic skills are first taught in the part-task trainer and the normal driving simulator. Then, participants take part actively in road traffic, with certain restrictions, e.g. as concerns speed. Based on this driving experience, the next level follows, with additional courses, such as targeted behavioral training in emergencies, using correspondingly highperformance full-task simulators. The procedure can be extended for certain groups of drivers, e.g. for those with little experience in driving or when hazardous goods drivers have special requirements. Refresher lessons are repeated, so that further improvements can be achieved by providing feedback, or to curb undesirable developments. One further word about costs. Driving simulators cannot – and should not – replace vehicles. They complement and alter teaching and training concepts. One-dimensional monetary assessments of future simulator courses are limited and remain open to criticism, as environmental variables are transformed into fictitious sums of money, which are never actually paid. One example is the expected benefit to the environment due to primary energy savings: simulator manufacturers talk of savings of 50 percent. Other examples are changes and reductions in noise and exhaust gas pollution, especially in built-up areas, or reductions in the number of accidents if courses are increasingly transferred to simulators. This is all still speculation: the forecasts may be wrong. Indeed, in view of our constantly changing body of knowledge, these hypotheses cannot be expected to remain unchanged in future. For the sake of completeness, here is a final example of a cost calculation. The operator of the two German driving simulator centers reckoned with prime costs of 1,000 euros per hour of simulator use. For a one-day seminar with 18 participants, 1,500 euros was charged per professional driver a day – but was rarely obtained, as discounts need to be granted. It is a sobering thought that the average simulator driving time per participant at these seminars is only slightly under two hours per course. This supports the idea that a move is needed toward multi-dimensional thought, decision and actions. Biological and physical environmental phenomena such as noise and exhaust gas pollution, or improved quality of life, are, after all, multifunctional, dynamic and multidimensional (Holub & Tappeiner, 1993), and corresponding steps valuable in themselves. This insight should not be masked by simple monetary considerations and one-dimensional sums, even if these figures, stamped with the hallmark of statistics, can be better marketed and hopefully add to the policy decisions and specification agreements required for the technological genesis of


133

driving simulation. Instead, it is also intended to lead to necessary, overdue policy decisions and the establishment of wide-scope definitions of aims, application and features. With these, after all, the technological genesis of driving simulation can achieve success in the long term.

Hardships of the best “What are you working on?” Mr. Keuner was asked. Mr. K. replied: “I’m having a hard time, I’m preparing my next error.” – Bert Brecht (1933)

Chapter 6

Bibliography

American Educational Research Association (AERA), American Psychological Association (APA) & National Council on Measurement in Education NCME) (1985) Standards for Educational and Psychological Testing. Washington DC: American Psychological Association Anastasi A (1990) Psychological Testing. New York: Macmillan Aschenbrenner KM, Biehl B & Wurm GW (1976) Mehr Verkehrssicherheit durch bessere Technik? Felduntersuchung zur Risikokompensation am Beispiel Antiblockiersystem ABS. Bergisch Gladbach: Bundesanstalt für Straßenwesen p 246 Bartlett MS (1947) Multivariate Analysis. J Royal Statistical Society Series B, 9 Suite 176–197 Beelitz N, Frost S, Käppler WD & Weber H (1995) Trainingsprogramm für Gefahrgutfahrer – Drehbuch fahrpraktisches Training. Bonn: Käppler & Partner Report 6.11 Beelitz N, Käppler WD & Mehl K (1995) Trainingsprogramm für Gefahrgutfahrer – Handout zu Train the Trainer. Bonn: Käppler & Partner Report 8 Benczik A & Spoerer E (1978) Bewährung im Straßenverkehr. Faktor Mensch im Verkehr, Darmstadt, 29 Bergius R (1964) Lernen und Denken. Handbuch der Psychologie. Vol 1/2. Göttingen: Hogrefe Biehl B (1976) Simulatoren – ein Fortschritt für die Erforschung des Verkehrsverhaltens? Verkehrssicherheit 22:1, pp 3–6 Biehl B & Brown ID (1986) A Comparison of Driver Training in the Federal Republic of Germany and Great Britain. Bergisch Gladbach: Bundesanstalt für Straßenwesen, Forschungsprojekt 8507 Binninger E (2000) Simulation Technology. Lectures Technical University of Aachen (in German, unpublished) Birkenbihl M (1995)Train the Trainer: Arbeitshandbuch für Ausbilder und Dozenten. 12 Ed. Landsberg/Lech: Verlag Moderne Industrie Blaauw GJ (1982) Driving Experience and Task Demands in Simulator and Instrumented Car: A Validation Study. Human Factors 24:4, pp 473–486 Bortz J (1977) Lehrbuch der Statistik für Sozialwissenschaftler. Berlin: Springer Braun CC, Bowers CA. & Holmes BE (1993) Impact of Task Difficulty on the Acquisition of Aircrew Coordination Skills. In Proc 37th Annual Meeting Human Factors and Ergonomics Society. Santa Monica. Ca. : Human Factors and Ergonomics Society, pp 1262–1266 Brecht B (1967/2001) Stories of Mr. Keuner. Translated by Martin Chalmers. City Lights Books, San Francisco Brehmer B Gregersen NP & Morén B (2–4 May 1991) Group Methods in Safety Work. In Workshop on New Technology and Work. Control of Risk. Bad Homburg von Bressensdorf G, Heilig B, Heinrich HC, Käppler WD & Weinand M (1995) Eignung von Pkw-Fahrsimulatoren für Fahrausbildung und Fahrerlaubnisprüfung. Bergisch Gladbach: Bundesanstalt für Straßenwesen Report M 50

135

136

6 Bibliography

Brown ID, Groeger JA & Biehl B (1987) Is Driver Training Contributing Enough Towards Traffic Safety? In Road Users and Traffic Safety. (pp 135–156, Eds Rothengatter JA & de Bruin RA) Maastricht: Van Gorcum Büschgens G (1993) Verkehrssicherheit als soziales und soziologisches Problem. Verkehrssicherheit 39 pp 150–156 Die Fahrschüler-Ausbildungsordnung (2004). Berlin: BGBl. I S. 2092 Dieterich R, Schmid I, Ritscher H & Hansmann KW (1993) Lkw-Simulatoren als Ausbildungsmittel. Psychologische, pädagogische, technische und o¨ konomische Studie zur Eignung von Lkw-Fahrsimulatoren in der Ausbildung zum Führerschein CE. Hamburg: Universität der Bundeswehr Dorsch F (1987/1991) Psychologisches Wörterbuch. Bern Stuttgart Toronto: Huber 11th edn Drosdol J & Panik F (1985) The Daimler Benz Driving Simulator. A Tool for Vehicle Development. Detroit Mich: SAE–Technical Paper Series 850334 Duncan J, Williams P & Brown ID (1991) Components of Driving Skill: Experience Does Not Mean Expertise. Ergonomics 34:7, pp 919–937 Edwards DS, Hahn CP & Fleishman EA (1977) Evaluation of Laboratory Methods for the Study of Driver Behaviour: Relations between Simulator and Street Performance. J Applied Psychology 62:5, pp 559–566 Evans L (1991) Traffic Safety and the Driver. New York: Van Nostrand Reinhold Fastenmeier W (1988a) Ein Katalog von Verkehrssituationen für repräsentative Fahrttypen. Munich: Technische Universität, research report, September Fastenmeier W (1988b) Vor- und Nachteile von Kfz-Simulatoren. Munich: Technische Universität, research report, November Fietkau HJ & Timp DW (1989) Einstellungen und Kognitionen gegenüber Umweltrisiken im beruflichen Alltag. Bericht u¨ ber eine empirische Untersuchung bei Fahrern im Gefahrguttransport. WZB Berlin, FS II 89–308 Flexman RE & Stark EA (1987) Training Simulators. In Handbook of Human Factors (pp 1012– 1038, ed Salvendy) New York: Wiley Florian M (1994) Highway–Helden in Not. Berlin: Edition Sigma Förstberg, J (2000) Ride comfort and motion sickness in tilting trains. Doctoral thesis. Stockholm: Department of Vehicle Engineering, Royal Institute of Technology. TRTITA-FKT Report 2000: 28 Forschungsgesellschaft für Straßen- und Verkehrswesen (1984) Richtlinien für die Anlage von Straßen RAS. Teil: Linienführung (RAS-L). Abschnitt 1: Elemente der Linienführung RAS-L-1. Bonn: Kirschbaum Frieling E & Hoyos C (1978) Graf Fragebogen zur Arbeitsanalyse – FAA. Bern: Huber Frieling E & Sonntag K (1987) Lehrbuch Arbeitspsychologie. Bern: Huber Frost S & Käppler WD (1995a) Trainingsprogramm für Gefahrgutfahrer–Tätigkeitsbeschreibungen von Gefahrgutfahrten. Bonn: Käppler & Partner Report 5 Frost S & Käppler WD (1995b) Trainingsprogramm für Gefahrgutfahrer–Tätigkeitsbeschreibungen von Stückgutfahrten. Bonn: Käppler & Partner Report 7 Geißler KA (1992) Zeit leben. Weinheim Berlin: Quadriga 4th Ed Gibson JJ (1982) Wahrnehmung und Umwelt. München: Urban & Schwarzenberg Godthelp J (1992) Tasks, activities and automation when driving a car. Personal communication with author. Godthelp J & Käppler WD (1998) The Effects of Vehicle Handling Characteristics on Driving Strategy. Human Factors 30 pp 219–229 Goettl BP (1993) Analysis of Skills on a Flight Simulator. In Proc 37th Annual Meeting Human Factors and Ergonomics Society, Santa Monica. Ca. : Human Factors and Ergonomics Society, pp 1257–1261 Gopher D Weil M & Siegel D (1989) Practice Under Changing Priorities: An Approach to the Training of Complex Skills. Acta Psychologica. 71 pp 147–177 Gubser F & Spörli S (1970) Teilsimulation in der verkehrspsychologischen Diagnostik. NZZ 347 Guilford JP (1967) The Nature of Human Intelligence. New York: Mc Graw Hill

6 Bibliography

137

Guilford JP (1977) Way Beyond the IQ. New York: Creative Education Foundation Guilford JP (1981) Higher Order Structure of Intellect Abilities. Multivariate Behavioural Research 16 411–435 Guilford JP (1986) Creative Talents. Buffalo, New York: Bearly Limited Hacker W (1986) Arbeitspsychologie. Bern: Huber Häcker H (1972) Die Umweltsimulation als Forschungsmethode der Verkehrspsychologie. Fahreignung und Verkehrssicherheit 20 pp 13–22 Hale AR & Glendon AI (1987) Individual Behaviour in the Control of Danger. Amsterdam: Elsevier Hale AR & Singer MJ (1989) Simulation Fidelity in Training System Design. New York: Springer Harvey C (1985) The E-Type. End of an Era. Sparkford: Oxford Illustrated Press Haug EJ (1990) Feasibility Study and Conceptual Design of a National Advanced Driving Simulator. Iowa City: University of Iowa College of Engineering Herrmann T, Hofstätter P, Huber HP & Weinert FE (1977) Handbuch psychologischer Grundbegriffe. Munich: Kösel Hoffmann J, Kammerhoff H & Käppler WD (1994) Trainingsprogramm für Gefahrgutfahrer – ¨ Ubersicht, Analyse und Bewertung der Literatur zu Simulatorausbildung und -Forschung, Berufskraftfahrerausbildung, Unfallursachen, Verkehrssicherheitsforschung und Sicherheitstrainings. Bonn: Käppler & Partner Report 1 Hollnagel E (1981) Verification and Validation in the Experimental Evaluation of New Designs for Man-Machine Systems. Roskilde Denmark: Riso National Laboratory Report M 2300 Holte H (1994) Kenngrößen subjektiver Sicherheitsbewertung. Berichte der Bundesanstalt für Straßenwesen. Bremerhaven: Wirtschaftsverlag NW Holub HW & Tappeiner G (1993) Irreführende Bilanz. Zeit 49:49, pp 27 van der Horst ARA (1990) A Time-Based Analysis of Road User Behavior in Normal and Critical Encounters. Diss Soesterberg, NL: Practicum Hoskovec J (1972 unpublished) Experimentelle Forschung u¨ ber die Ausbildung der Lenker in der CSSR, Fahrsimulator Typ 65 und Gefahrensituationen. Paper 14 at ivv-Kongress, Vienna Hulbert S & Wojcik CJ (1960) Driving Simulator Research. In 39th Annual Meeting Highway Research Board, pp 1–13 Janssen WH (1988) Risk Compensation and the Effect of an Incentive: A Laboratory Study. Soesterberg Niederlande: TNO Human Factor Research Institute Report C–26 Jensch M, Spoerer E & Utzelmann HD (1978) Verkehrsverhaltenslehre. Bergisch Gladbach: Bundesanstalt für Straßenwesen, Bereich Unfallforschung Report 20 Johannsen G (1990) Fahrzeugführung. In Ingenieurpsychologie (pp 426–454, eds Hoyos G and Zimolong, B) Göttingen: Hogrefe Käppler WD (1991) Simulators in Safety Studies. In Proc 3rd European Workshop on Recent Developments in Road Safety Research (pp 41–43). Linköping Sweden: National Swedish Road and Traffic Institute, VTI Report 366A Käppler WD et al. (1992) Simulator Technology: Analysis of Applicability to Motor Vehicle Travel. Washington DC: National Research Council – Transportation Research Circular 388 Käppler WD (1993a) Views on the Role of Simulation in Driver Training. In Proc 12th European Annual Conference on Human Decision Making and Manual Control (pp 5–12 – 5–17, eds Johannsen G et al.). Kassel: Universität Gesamthochschule Käppler WD (1993b) Analyse und Bewertung der Ergebnisse von Erprobung und Truppenversuch des Vorhabens “Verbesserung der Tornado-Simulatoren Tiefflug”. Wachtberg: Forschungsinstitut für Anthropotechnik Report 102 Käppler WD (1993c) Beitrag zur Vorhersage von Einschätzungen des Fahrverhaltens. Diss Universität Kassel. Düsseldorf: VDI Fortschrittsberichte series 12 No. 198 Käppler WD (1994a) Fahrsimulatoren in der Fahrschulausbildung. In Optimaler Automatisierungsgrad von Mensch-Maschine-Systemen (pp 89–100, ed Gärtner KP). Bonn: Deutsche Gesellschaft für Luft- und Raumfahrt EV, Report 94–01

138

6 Bibliography

¨ Käppler WD (1994b) Uberlegungen zu Ausbildungssystemen für Berufskraftfahrer mit Fahrzeugen, Simulatoren und weiteren Medien. In Ausbildung mit Simulatoren (pp 211–225, ed Gärtner KP). Bonn: Deutsche Gesellschaft für Luft- und Raumfahrt e.V., Report 94–03 ¨ Käppler WD (1994c) Trainingsprogramm für Gefahrgutfahrer – Ubersicht, Analyse und Bewertung der Lernziele und -Inhalte sowie Konzeption des Programms. Bonn: Käppler & Partner Report 2 ¨ Käppler WD (1995a) Trainingsprogramm für Gefahrgutfahrer – Ubersicht der Literatur zur Trainingskonzeption. Bonn: Käppler & Partner Report 3 Käppler WD (1995b) Trainingsprogramm für Gefahrgutfahrer – Analyse und Bewertung der Literatur zur Trainingskonzeption und spezifische Belastungen der Arbeitstätigkeit von Gefahrgutfahrern. Bonn: Käppler & Partner Report 4 Käppler WD (1995c) Ausbildungs- und Trainingssysteme mit Simulatoren. Drehbuch Mineralöltransport. Bonn: Käppler & Partner Report 9.3 Käppler WD (1996a) Ausbildungs- und Trainingssysteme mit Simulatoren. Handbuch Anwendung. Bonn: Käppler & Partner Report 9.1 Käppler WD (1996b) Ausbildungs- und Trainingssysteme mit Simulatoren. Handbuch Grundlagen. Bonn: Käppler & Partner Report 9.2 Käppler WD (1996c) Briefing, Debriefing & Replay im Lufthansa Simulatortraining. Bonn: Käppler & Partner Report 10 Käppler WD (1996d) Bewegungssimulation Tanksattelzug – Zusammenfassung und Bewertung der Fahrversuche. Bonn: Käppler & Partner Report 11 Käppler WD (1996e) Tätigkeitsbeschreibungen von Linienbusfahrten. Bonn: Käppler & Partner Report 12 Käppler WD (1996f) Moderne Konzepte zur Polizei-Fahrausbildung. Bonn: Käppler & Partner Report 13 Käppler WD (1996g) Fahrschule 2000 – Konzeption einer zukünftigen Fahrschulausbildung. Bonn: Käppler & Partner Report 15 Käppler WD (1996h) Ausbildungs- und Trainingssysteme mit Simulatoren. Vergleichende Bewertung der Tätigkeitsanalysen. Bonn: Käppler & Partner Report 16 Käppler WD (1996i) Eignungs- und Einstellungstests mit Simulatoren – Verfahren und Messgütekriterien. Bonn: Käppler & Partner Report 18 Käppler WD (1997a) Human Factors. In Piloted Simulation in Low Altitude High Speed Mission Rehearsal (pp 76–85, ed Tomlinson BN). Neuilly F: NATO AGARD Advisory Report 333 Käppler WD (1997b) Verknüpfung fahrpraktischer Programme des DVR mit Simulatoren – Fahreranforderungsprofile. Bonn: Käppler & Partner Report 19 Käppler WD (1997c) Simulatordatenbasis für die Ausbildung von wirtschaftlicher Fahrweise. Bonn: Käppler & Partner Report 20 Käppler WD (1997d) Studie für die Fahrsimulation – Kenngrößen zur Beurteilung der Fahrweise. Bonn: Käppler & Partner Report 21 Käppler WD (1997e) Studie für die Fahrsimulation – Katalog Kritischer Situationen. Bonn: Käppler & Partner Report 22 Käppler WD (1997f) Studie für die Fahrsimulation – Ereignissteuerung und Featureliste. Bonn: Käppler & Partner Report 23 Käppler WD (2000) Verknüpfung fahrpraktischer Programme des DVR mit Simulatoren – Drehbücher Seltenes Ereignis & Handlungsfixierung. Bonn: Käppler & Partner Report 24 Käppler WD (2002) Untersuchung von Motorradunfällen. Bonn: Käppler & Partner Report 25 & Berufsgenossenschaften für den Einzelhandel Käppler WD (2004) Menschliche Fehler als Unfallursachen: Untersuchungen und Ergebnisse mit ARIADNE. (pp 197–212, ed Grandt, M) Bonn: Deutsche Gesellschaft für Luft- und Raumfahrt e.V., report 04–03 Käppler WD (2005) Untersuchung von Maßnahmen zur Verbesserung der Wahrnehmbarkeit motorisierter Zweiradfahrer im Straßenverkehr. Bonn: Berufsgenossenschaft für den Einzelhandel, Schaufenster Sicherheit

6 Bibliography

139

Käppler WD (2006) ARIADNE – Internet Based Error Risk Management and Prevention System. In Vanderhaegen F (ed) Proc 25th European Conference on Human Decision Making and Manual Control. Valenciennes, F. Käppler WD (2007) Das Sicherheitsmanagementsystem ARIADNE. Wachtberg: Report FKIEEFS/2007/145 ¨ Käppler WD & Alexander T (1995) Internationale Ubersicht Fahrsimulatoren. Wachtberg: Forschungsinstitut für Anthropotechnik Report 112 Käppler WD & Bernotat R (1985) Wirksystem Fahrer-Fahrzeug-Umwelt. In Verkehrssicherheit ¨ Rheinland und Wirksystem Fahrer-Fahrzeug-Umwelt (pp 13–44, ed Rompe K) Cologne: TUV Käppler WD & Dalinger E (2005) Derivation of Safety Measures through Human Error Risk Analyses. In CA. Brebbia et al. (eds) Safety and Security Engineering. Southampton, Boston: Witpress, pp 253–263 Käppler WD & Mehl K (1997) Trainingskonzepte zur Verbesserung der menschlichen Zuverlässigkeit am Beispiel Berufsfahrerschulung. In Menschliche Zuverlässigkeit, Beanspruchung und benutzerorientierte Automatisierung (pp 163–172, ed Gärtner KP). Bonn: Deutsche Gesellschaft für Luft- und Raumfahrt e.V., Report 97–02 Käppler WD & Mehl K (1999a) Neue Ansätze zur Verbesserung der Menschlichen Handlungszuverlässigkeit. In Ergonomische Gestaltungswerkzeuge in der Fahrzeug- und Prozessführung (pp 13–23, ed Gärtner KP). Bonn: Deutsche Gesellschaft für Luft- und Raumfahrt e.V., report 99–02 Käppler WD & Mehl K (1999b) Simulator Based Training System Design – Recent Progress & Lessons Learned with Truck Driving. Bonn: Deutsche Gesellschaft für Luft- und Raumfahrt e.V., Report 99–02 Käppler WD & Motz F (1995) Entwurf eines integrierten Ausbildungssystems mit Fahrzeugen und Simulatoren für Militärkraftfahrer nach Stufenplan CE. Wachtberg: Forschungsinstitut für Anthropotechnik Report 107 Käppler WD, Preßler R & Specht D (2008) Sicherheitsmanagement mit ARIADNE. In Schmidt, Schlick, Grosche (eds) Ergonomie und Mensch-Maschine-Systeme. Heidelberg: Springer, pp 437–462 Käppler WD & Voß M (1988) Trainingssimulatoren in der Aus- und Weiterbildung von Autofahrern aus Expertensicht. Wachtberg: Forschungsinstitut für Anthropotechnik Report 78 Kaiser HF & Dickman K (1959) Analytic Determination of Common Factors. Amer Psychol 14 pp 425 ff Katz D & Stotland EA (1959) Preliminary Statement to a Theory of Attitude Structure and Change. In A Study of a Science (ed Koch S) Vol 3 Formulations of the Person and the Social Context. New York: McGraw Hill Kelley CR (1969) The Measurement of Tracking Proficiency. Human Factors 11:1, pp 43–64 Kennedy RS, Berbaum KS & Smith MG (1993) Methods for Correlating Visual Scene Elements with Simulator Sickness Incidence. In Proc 37th Annual Meeting Human Factors and Ergonomics Society, Santa Monica. Ca. : Human Factors and Ergonomics Society, pp 1252–1256 Kirwan B & Ainsworth LK (1993) A Guide to Task Analysis. London: Taylor & Francis Knie A (1994) Wankel-Mut in der Autoindustrie. Berlin: Edition Sigma Korteling JE (1991) Visuele informatie in rijsimulatoren: literaturstudie en onderzoeksvoorstel (Visual Information in Driving Simulators: Literature Study and Research Proposal). Soesterberg Niederlande: TNO Human Factor Research Institute Report IZF B–5 Küting HJ (1986) Verkehrserziehung in der Primarstufe: Sozial-kognitive Anforderungen und Konzeption. Bergisch Gladbach: Bundesanstalt für Straßenwesen Report 146 Leontiev AN (1973) Probleme bei der Entwicklung des Psychischen. Frankfurt: Athenäum Lindgren HC (1973) Einführung in die Sozialpsychologie. Weinheim: Beltz Lotens W, Allender L, , Belyavin A, Cain B, Castor M, Käppler WD, Lundin M, Thomas-Meyers G & Wallin N (2005) HFM–128 NATO Research Task Group on Representation of Human Behaviour in Constructive Simulation. In Proceedings BRIMS Conference. Los Angeles Mehl K & Käppler WD (1998a) Simulation – Bedeutung und Anforderungen sowie Bemerkungen zur konzeptuellen Logik der Systeme. In Anthropotechnik – gestern – heute – morgen

140

6 Bibliography

(pp 131–137, Ed Gärtner KP). Bonn: Deutsche Gesellschaft für Luft- und Raumfahrt e.V., Report 98–02 Mehl K & Käppler WD (1998b) Einsatz von Fahr- und Verkehrssimulation in der Berufsfahrerschulung. In Psychologie der Arbeitssicherheit (pp 98–102, eds Benda Hv & Bratge D) Heidelberg: Asanger Meyers Großes Taschenlexikon (1987). Mannheim Wien Zürich: BI Taschenbuchverlag Minsky M (1988) The Society of Mind. New York: Simon & Schuster Misumi J (1978) The Effects of Organizational Climate Variables, Particularly Leadership Variable and Group Decision on Accident Prevention. In 19th International Congress of Applied Psychology. München Misumi J (1982) Action Research on Group Decision Making and Organizational Development. In Social Psychology (eds H Hiebsch, H Brandstetter & HH Kelly, pp 94–98) Berlin: VEB Deutscher Verlag der Wissenschaften Müller A (1994) Was ein Fahrschüler alles lernt. Verkehrssicherheit 40:3, pp 102–106 Müller GE (1911) Zur Analyse der Gedächtnistätigkeit und des Vorstellungsverlaufs. Erg. Z. Ps. 5 Nissen MJ & Bullemer P (1987) Attentional Requirements of Learning. Evidence from Performance Measures. In Proc 6th International Symposium on Aviation Psychology. Columbus: Ohio State University, pp 272–277 Olsen RA (1973) Skid Simulator for Use in Driver Training and Research. In Highway Research Record 464, Driver Behaviour Research Methods. Washington DC: Highway Research Board Padmos P & Milders MV (1992) Quality Criteria for Simulator Images: A Literature Review. Human Factors 34:6, pp 727–748 Pain RF, Whittenburg JA, McBride RS & Baker GL (1973) Evaluation and Development of Driver Training. In Proc First International Conference on Driver Behaviour. Zurich Parker D, Reason JT, Manstead AS R & Stradling SG (1995) Driving Errors, Driving Violations and Accident Involvement. Ergonomics 38:5, pp 1036–1048 Patrick J (1992) Training: Research and Practice. London: Academic Press Pavlov IP (1926) Die höchste Nerventätigkeit (das Verhalten) von Tieren. Saarbrücken: VDM Müller. Reprint 3 ed München: Bergmann Pelz DC & Krupkat E (1974) Caution Profile and Driving Record of Undergraduate Males. Accident Analysis and Prevention 6 pp 45–58 Petsching M (1967) Der kleine Stowasser. Munich: G Freytag Pirandello L (1999) So ist es (wenn es Ihnen so scheint). Berlin: Propyläen Poulton EC (1971) Skilled Performance and Stress. In Psychology at Work (55–75, 1st edn, ed Warr PB). Harmondsworth: Penguin Books. Rasmussen J & Jensen A (1974) Mental Procedures in Real-Life Tasks: A Case Study of Electronic Trouble Shooting. Ergonomics 17:4, pp 293–307 Rasmussen J (1983) Skills, Rules and Knowledge; Signals, Signs and Symbols and Other Distinctions in Human Performance Models. IEEE Transactions on Systems, Man and Cybernetics, SMC–13:3, pp 257–266 Reason J. (1990) Human Error. Cambridge University Press Roscoe SN & Williges BH (1980) Measurement of Transfer of Training. In Aviation Psychology. Ames Iowa: Iowa State University Press Roscoe SN, Jensen RS & Gawron VJ (1980) Introduction to Training Systems. In Aviation Psychology. Ames Iowa: Iowa State University Press Rößger E, Zehle H & Wernicke J (1962) Vorgang und Terminologie der Raumfahrzeugführung. Luftfahrtechnik 8:8, p 208 Rüter G (1983) Sicherheitsorientierte Ausbildung von Berufskraftfahrern. Bergisch-Gladbach: Bundesanstalt für Straßenwesen Report 81 Salvendy G (2006) Handbook of Human Factors and Ergonomics. Hoboken: John Wiley & Sons 3rd Ed Schagen I van, Rothengatter JA & Brookhuis KA (1987) Formulating Educational Objectives for Young Cyclists and Moped Riders. In Proc Canadian Multidisciplinary Road Safety Conference V. Calgary, pp 166–173

6 Bibliography

141

Seidemann O (1976) Die Fahrprobe am Fahrsimulator – ein Verfahren zur Beurteilung der Verkehrseignung von Kraftfahrern. Verkehrssicherheit 22:1, pp 28–33 Senders JW & Moray N (1991) Human Error. Hillsdale New Jersey: Lawrence Erlbaum Associates Seydel U & Beetz E (1978) Wirksamkeitsanalyse zum Sicherheitstraining des Deutschen Verkehrssicherheitsrates. Verkehrssicherheit 24:3, pp 113–119 Sonntag K (1989) Trainingsforschung in der Arbeitspsychologie. Bern Stuttgart Toronto: Huber Springer J, Göbel M, Hedicke V & Rötting M (1993) Einflüsse taktilen Feedbacks bei primär visuellen Trackingaufgaben. In Virtual Reality. Berlin: Springer Summala H (1989 unpublished) The Human Factor in Highway Safety. Recent Results on “Soft” Countermeasures Statistisches Bundesamt (2007) Verkehrsunfälle Zeitreihen. Wiedsbaden: Statistisches Bundesamt. Steininger K (1995) Ausbildung und Training. In Personalentwicklung für komplexe MenschMaschine-Systeme (eds Steininger K, Fischbauer S & Goeters KM) Weinheim: Beltz/PVU Thorndike EL & Gates Al (1930) Elementary Principles of Education. New York: MacMillan Tränkle U, Gelau C & Metker T (1988) Situationsbezogene Risikoeinschätzung junger Fahrer. Bergisch Gladbach: Bundesanstalt für Straßenwesen Projekt 8307/3 Tschernitschek E (1978) Verkehrssicherheitsprogramm eines Vertriebsunternehmens. Berufsgenossenschaft 2 Utzelmann HD (1985) Evaluation der Kurse für mehrfach auffällige Kraftfahrer. Bremerhaven: Wirtschaftsverlag NW de Vos A, Godthelp J & Käppler WD (1999) Subjective and Objective Assessment of Manual Supported and Automated Vehicle Control. In Vehicle Performance: Understanding Human Monitoring and Assessment (pp 97–120, ed Pauwelussen JP). Lisse, Netherlands: Swets & Zeitlinger Winkler W (1988) Wirksamkeit von Kursen für wiederholt alkoholauffällige Kraftfahrer. Bergisch Gladbach: Bundesanstalt für Straßenwesen, Unfallforschung Report 64 Wittig W (1994) Integration eines psychologischen Trainings in die Simulatorausbildung zukünftiger Schiffsoffiziere. In Ausbildung in Simulatoren (pp 157–167, ed Gärtner KP). Bonn: Deutsche Gesellschaft für Luft- und Raumfahrt e.V., DGLR Report 94–03

Smart Driver Training Simulation: Save Money. Prevent

Save Energy, Save Money

Save Energy, Save Money

How to Save Money

Train Smart!, Precision Training

Driver

Smart Money - June 2011

Smart Money - March 2011

Money Sense- Getting Smart with Your Money

Money Sense Getting Smart With Your Money

Training for a Smart Workforce

Driver

Driver

Smart Women, Smart Money: Live the Life You Want

Star Driver

Money Girl's Smart Moves to Grow Rich

The Driver

Star Driver

Human Factors in Simulation and Training

The Driver

The Driver

Shop Smart, Save More: Learn The Grocery Game and Save Hundreds of Dollars a Month

Smart

Consumer Guide to Home Energy Savings: Save Money, Save the Earth

The Truck Driver

Using Divorce Mediation: Save Your Money & Your Sanity

Money, Money, Money

Money, Money, Money

Money, Money, Money

Money, Money, Money

Smart Driver Training Simulation: Save Money. Prevent

Save Energy, Save Money

Save Energy, Save Money

How to Save Money

Train Smart!, Precision Training

Driver

Smart Money - June 2011

Smart Money - March 2011

Money Sense- Getting Smart with Your Money

Money Sense Getting Smart With Your Money

Training for a Smart Workforce

Driver

Driver

Smart Women, Smart Money: Live the Life You Want

Star Driver

Money Girl's Smart Moves to Grow Rich

The Driver

Star Driver

Human Factors in Simulation and Training

The Driver

The Driver

Shop Smart, Save More: Learn The Grocery Game and Save Hundreds of Dollars a Month

Smart

Consumer Guide to Home Energy Savings: Save Money, Save the Earth

The Truck Driver

Using Divorce Mediation: Save Your Money & Your Sanity

Money, Money, Money

Money, Money, Money

Money, Money, Money

Money, Money, Money

Recommend Documents