Generic Intelligent Driver Support

J A Michon

• Publisher: CRC Press
• Year: 2021
• Print ISBN:
  9780748400690
• eBook ISBN:
  9781000447569
• Topic: Technology & Engineering
• Subtopic: Computer Science General
• Index: Technology & Engineering

This book summarizes the activities of the Generic Intelligent Driver Support (GIDS) Consortium and offers recommendations for successful GIDS implementation. It is based on the GIDS Project, a part of the EC-funded Dedicated Road Infrastructure for Vehicle Safety in Europe Programme.

Part I
The GIDS concept

Chapter 1
Introduction: a guide to GIDS

John A. Michion, Alison Smiley

1.0 Chapter outline

In this introductory chapter an overview is presented of the nature and development of GIDS, the Generic Intelligent Driver Support system. The development of GIDS took place between 1989 and the middle of 1992, as part of the EEC DRIVE programme. The present volume, summarizing the activities of the GIDS Consortium, was prepared at the conclusion of the project.

An overview of an intelligent, or adaptive, driver support system is presented in Section 1.1. Next the objectives (Section 1.2) and some historical background (Section 1.3) of the GIDS project are provided. The position of GIDS within the larger framework of Road Transport Informatics (RTI) is reviewed in Section 1.4.

The next four sections present the GIDS perspective on driver support. In Section 1.5 we outline the domain that is covered by the GIDS concept and the constraints that had to he imposed to keep the project within practical limits. In Section 1.6 we review the functional characteristics of the GIDS system, and Section 1.7 is devoted to the way in which GIDS has been implemented.

Finally, in Section 1.8 a brief preview of the subsequent chapters of this book is presented.

1.1 What is GIDS?

Introduction

The overall objective of the GIDS project has been "to determine the requirements and design standards for a class of intelligent co-driver (GIDS) systems that are maximally consistent with the information requirements and performance capabilities of the human driver" (from the GIDS project proposal, GIDS, 1988). The project has resulted in recommendations for such systems and in a (limited) operational prototype demonstrating the essential features of the GIDS concept, both under simulated and real world traffic conditions. The GIDS project - officially known as DRIVE Project VI041 "Generic Intelligent Driver Support (GIDS)" was part of the DRIVE Programme which was initiated in 1988 by the Commission of the European Communities to stimulate and coordinate the introduction of modern Road Transport Informatics (RTI). DRIVE stands for Dedicated Road Infrastructure for Vehicle safety in Europe.

Co-driver systems or driver support systems - the latter being the term to be used throughout this book - derive their usefulness to a large extent from the fact that vehicle operators must cope with a growing amount of information of an increasingly complex nature. This is caused by several factors, including increasing traffic density, an increasing number of on-board and roadside sources of information and, last but not least, by the increasing amount of additional in-vehicle equipment, such as telephones and fax machines. A driver support system, such as GIDS, will help to counter the information pollution that is threatening the vehicle operator, by filtering, interpreting, integrating, prioritizing, and presenting the information from any number of sensors and applications.

This avalanche of information - much of which will eventually be generated by RTI systems resulting from the DRIVE programme - is likely to have an impact on almost every aspect of the driving task. It will affect route planning as well as navigation, manoeuvring, and elementary vehicle control. Unless regulatory action is taken this information will eventually be presented to the driver in an essentially incoherent fashion, irrespective of its importance or appropriateness. A critical function of GIDS, as of any other driver support system, is to protect the driver from being overwhelmed by such uncoordinated information.

The innovative feature of GIDS is that it is the first system ever to take into account, in an adaptive fashion, (some of) the intentions, capabilities, and limitations of the individual driver. Driver support systems should enable drivers to cope with the driving task more easily, safely and efficiently (and, indirectly, at diminished cost to the environment).

By meeting its overall objective, the GIDS project will help to further the goals of the DRIVE programme which were: "to increase traffic safety, to improve transport efficiency, to reduce energy consumption and to improve the environment" (as stated in the DRIVE Call for Proposals, DRIVE, 1988).

A scenario

To obtain a feeling for what GIDS might eventually become, imagine you are travelling to Switzerland for a holiday in the Alps. Upon arrival your rental car is waiting for you at the airport. In order to get going, you simply insert your personalized GIDS smart card in the receptacle on the dashboard. Instantly the vehicle will recognize who you are and adapt automatically to your individual needs and traits. It will know, say, that you have never driven on mountain roads before, and that you have a tendency to brake very briskly. Knowing these and other things, the GIDS system will be prepared to guide you slowly through hairpins and keep you far from those nasty 15⁺ per cent descending stretches of mountain road, at least for the first couple of days. Moreover, realizing that your knowledge of, say, French and Italian is marginal, it will translate all the local traffic information into your native language, including whatever there is to read on the various road signs.

The G, I, D, and S of GIDS

Whilst admitting that this is a somewhat far-fetched sketch of the performance of GIDS, let us now consider what GIDS is by looking more closely at the constituent concepts of the term.

The term generic refers to the fact that GIDS has been conceptualized in such a way that it may increase in scope and complexity with the development of new technological capabilities. The system as it figures in this volume covers a limited but realistic set of tasks, namely such tasks as can presently be defined unambiguously and for which equipment is currently feasible. GIDS is based on a communication protocol, that is, an agreed way of communication between the driver and the system. As a result the GIDS system and the external inputs to the system can be extended to accommodate subtasks and information sources that may be added later. In other words, GIDS is generic in the sense that it is not specific to the automation of particular driving tasks or rigidly defined categories of information.

GIDS contains intelligence of a kind that other current support systems do not. Ideally a GIDS system should be able to manage the stream of information to the driver in accordance with the driving task, the driving conditions, and the state the driver is in. Current support systems lack this type of intelligence. They do not select or integrate information according to the demands of the driving task or according to the state and intentions of the driver. Not only do conventional systems not select the appropriate level of detail, they also do not prevent the driver from selecting a level of detail that is inappropriate in the face of the current situation. With currently available technology drivers can be using a navigation system, talking on a cellular phone, and checking on their moment-to-moment fuel economy. The information from these multiple sources is not integrated in conventional systems and therefore unsafe situations may arise, due to overload or contradictory advice. GIDS is designed to prevent drivers from overloading themselves because it will integrate the available information from various sources, taking driver needs and intentions into account.

In designing GIDS to support a variety of driving tasks, the physical, perceptual and cognitive characteristics of the driver needed consideration. This includes his or her physical capacities, perceptual capacities, and cognitive capacities. An important additional characteristic of the driver which, nevertheless, has not been explicitly considered in the GIDS project, is the driver's emotional state. This deliberate omission has been an occasional source of criticism. After all, emotion is a powerful determinant of driving behaviour: a driver who is preoccupied with some stressful personal situation may be inattentive, and a driver who is feeling aggressive may speed, or overtake unsafely, and so on. However, only those aspects of emotional and motivational states that are observable, such as higher speeds, may be considered in a GIDS system, but thus far these aspects do not, unfortunately, provide unambiguous cues about the driver's state.

GIDS provides support at three major levels defining the driving task: navigation, manoeuvring and control. Besides support at each of these levels, the GIDS system provides support on a "meta-level". This is the area in which the GIDS concept differs most from conventional support systems.

First, information from individual support systems serving each of the three task levels is integrated and prioritized so that the system can respond appropriately to the driver's current situation. This means that the mental workload of the driver is taken into account in the presentation of information.

Second, the GIDS system can provide adaptive feedback. This includes evaluation of prior performance, instructive feedback, and progressive modification of support structure. The latter refers to such possibilities as the system gradually giving less and less route information as the driver has more experience of a particular route, down to the point where only information about changed conditions, such as a new construction zone, is presented.

1.2 Practical objectives and core activities

The GIDS overall objective, stated in Section 1.1, is too abstract to offer much practical guidance. In more concrete terms, the GIDS consortium has pursued the following set of practical goals, as stated in the original GIDS proposal:

• — define detailed functional requirements of generic intelligent driver support (GIDS) systems;
• — determine the impact of new road transport informatics (RT1) systems on the task representations and behaviours of drivers with respect to navigation, manoeuvring, and control aspects of the driving task;
• — determine the interactive communication (display and dialogue) between the driver and the new RTI systems, inclusive of adaptive feedback;
• - develop the required hardware and software that will lead to the implementation of a prototypical GIDS system. This prototype shall incorporate the substantive core of the GIDS concept;
• — determine the impact of systems that meet the GIDS specifications on driving safety, efficiency, training, and system acceptance;
• — demonstrate the validity of the GIDS concept in field tests.

These objectives have been pursued in two major stages. The first stage roughly covered the first 18 months of the project. It dealt with the specification of behavioural requirements and technical characteristics for a GIDS prototype of limited functionality. The second stage, initially intended to cover the period 1 July 1990 until 31 December 1991, but ultimately lasting until the middle of 1992, was devoted to the implementation and evaluation of this prototype.

We distinguish four major component activities that, together, have enabled the GIDS consortium to achieve the stated objectives according to plan.

The first activity consisted of a definition of the basic functional and operational requirements of GIDS systems. For this purpose several basic functional domains were distinguished. Altogether five such functional domains have been recognized as basically covering the whole domain of driving tasks: planning, manoeuvring, control, adaptive feedback instructions, and functional integration of non-driving activities, such as carrying on a telephone conversation while driving. Within each domain a characteristic application was then selected, allowing us to keep the size and complexity of the prototype within reasonable bounds whilst retaining the essentially generic nature of the GIDS concept.

The second activity was a critical examination of the functional and operational requirements of GIDS systems. For this purpose a series of preliminary studies literature reviews and pilot experiments - were carried out early in the project. The results of these studies initially gave rise to a working definition of the GIDS concept and later to operational recommendations for GIDS systems in general and for the GIDS prototype in particular

The results of the component studies were integrated into concrete design specifications, incorporating the substantive core of GIDS system functions and operations. These specifications guided the construction of the prototype GIDS system. Under the terms of the present project a prototype of limited functionality has become available, in two versions. The first is implemented in a genuine automobile for on-the-road studies, the second is part of a simulation fac...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Contents
• Project Colophon
• List of Contributors
• Preface
• Part I: The GIDS concept
• Part II: The GIDS system
• Part III: Performance and perspective
• Author index
• Subject index