The DETER (Detection, Enforcement & Tutoring for Error Reduction) project is concerned with the development and testing of prototypes of on-site, in-vehicle and integrated monitoring, tutoring and enforcement systems. On the one hand, the project aimed to accomplish this goal by carrying out simulator and field studies of in-vehicle and integrated monitoring, tutoring and enforcement systems to assess reliability, validity and behavioural effects. On the other hand, field studies were carried out to assess user acceptance, legal-administrative scope and limitations of on-site tutoring and enforcement systems within the framework of a pilot project implementation. The on-site part of the project started from an existing system in Norway which was developed for on-site detection of road user errors and violations of special interest, i.e. for reasons of severity and frequency. The sensors and infrastructure developed were adapted according to the objectives of the project, i.e. identifying offenders and recognizing vehicles for feedback and enforcement purposes. Classes of driving errors and violations which could be detected and thereupon drivers punished using these automatic on-site systems were identified. The technologies which comprise such systems and the ways they can be used to support enforcement were determined and described. The effects of different types of feedback indicated that more precise evidence was required before on-site systems were able to carry messages which relate only to the specific behaviour of an individual driver. Therefore, as a first step, it was suggested that the effects of collective feedback messages (i.e. information about what drivers in general (should) do at a particular site), was to be assessed using the proposed on-site systems. It could then be determined whether these effects could be extended by using combinations of collective messages and different levels of enforcement, and different contingencies between the behaviour of individual drivers and the presentation of different types of collective messages. Driving errors and violations were also determined by an in-vehicle subsystem `behaviour comparator'. The work on this in-vehicle behaviour assessment system concerned the development of a prototype in-vehicle behaviour assessment system that is capable of detecting deviations from the required behaviour by comparing the behaviour registered by the in-vehicle detection system with the (local) normative behaviour as contained in an adaptive database, updated by a site-vehicle communication system. The prototype system is a sort of expert-system that runs on any Personal Computer system. In the comparator six types of normative behavioural violations can be considered, each requiring its own specific set of in-vehicle and externally delivered data:(i) speed violations, i.e. driving faster than the local speed limit; (ii) overtaking where not allowed; (iii) headway violations, i.e. too close following of a car-in-front; (iv) neglect stop sign, i.e. passing STOP sign without actually stopping; (v) neglect traffic lights, i.e. ignoring red light; and (vi) use of restricted lanes. Reduced driver vigilance is to be determined by a subsystem `driver impairment monitor' (DIM). This subsystem was designed to detect those deleterious changes in driver behaviour resulting from the influence of fatigue, stress and possibly alcohol and drugs. The DIM was divided into three functional units, the vehicle sensors that collect instantaneous driving data, the expert system that analyses and interprets these data and the storage/retrieval device which is used as a template of normal driving behaviour. Prototypes of integrated behavioural assessment systems have been tested in driving simulators and in actual vehicles. On-the-road tests of Driver Impairment Monitoring (DIM) systems showed that changes in driver state as measured by physiology (EEG) coincide with changes in steering performance and subjective ratings, while auditory warnings with respect to impaired driving performance were followed by increased driver effort and steering performance. Positive effects of feedback regarding distance keeping to a car-in-front were found in three on-the-road studies. Time-headways below one second were significantly reduced in case of tutoring. It was found that subjects have difficulty to estimate inter-vehicle distances expressed in time headway. Explicit recommendations to the driver with respect to appropriate time headway, e.g. by means of a bar that changes colour, is to be preferred. In a final on-the-road test of the integrated system, tag-vehicle communication was found to function at all speeds. Tutoring messages of the tutoring and enforcement system were found to have a beneficial effect on law-compliance. Most drivers accepted these systems and considered them useful, although pleasantness-ratings were slightly negative. The factual introduction of in-vehicle enforcement systems has to deal with strict requirements of legitimacy of measurements. (A)
Abstract