Vassilis Papakostopoulos

Modelling Driver Behaviour in European Union and International Projects

Human (or operator) modelling has been an extensive area of research in many application areas, such as artificial intelligence, aviation, probabilistic risk assessments, system safety analysis and human performances in working contexts (Cacciabue et al., 1993; Baron et al., 1980). Still, human behaviour is fairly contextual and substantially different from one person to another. Thus, the initial linear models have been gradually replaced by nonlinear and even probabilistic models, based upon artificial intelligence (AI) principles, such as artificial neural networks or genetic algorithms. This becomes even more intrigued if we consider a complex behavioural task such as vehicle driving.

The effects of changes in the traffic scene during overtaking

Overtaking maneuvers were studied in real traffic, by comparing cases where a change in the opposite traffic occurred during the overtaking maneuver i.e., appearance of an oncoming car, with cases where no change occurred during the maneuver i.e., either an already apparent oncoming car or no oncoming car. In total 45 naturally occurring cases of overtaking were analysed. By examining the time headways (TH) between the overtaking car and the other cars involved, at the end of the maneuver, a significant correlation was found between the TH to opposite traffic and the TH rear to the overtaken car. In cases where a change in the opposite traffic occurred and the proximity to opposite traffic was below a critical TH (<4s) the appearance of a new-oncoming car caused the overtakers to actively press, both longitudinally and laterally, close to the overtaken car. In cases of overtaking where no change in the opposite traffic occurred, the TH rear to the overtaking vehicle was unaffected by the TH to the opposite traffic. Interestingly, in comparing cases where an oncoming car appeared during overtaking with cases where an oncoming car was apparent from the start of the maneuver both the TH to the opposite traffic and the duration of the maneuver were not found to be significantly different. It is suggested that drivers were probably expecting to be confronted with an oncoming car during the overtaking. However, the decreased available time to disambiguate this situation leads the overtaking driver to limit the rear safety margin of the vehicle being overtaken. The appropriateness of this practice, in terms of safety, remains questionable.

The “field of safe travel” revisited: interpreting driving behaviour performance through a holistic approach

ABSTRACT The present paper proposes a conceptual framework for the driver’s visual–spatial perceptual processes. Based on a theoretical analysis of driving proposed by Gibson and Crooks [(1938). A theoretical field-analysis of automobile-driving. The American Journal of Psychology, 51, 453–471. doi:10.2307/1416145], the developed field of safe travel (FoST) framework suggests that at any moment the driver constructs a “field” by integrating two perceptual entities: (i) the possible available spatial fields for locomotion and (ii) the driver’s mental image of ego-vehicle outer-line and motion dynamics. This framework is used to reinterpret in a unified way a number of disparate research findings reported in the literature concerning specific driving sub-tasks (e.g. lane keeping and car following). It is argued that the FoST framework may be used to predict drivers’ behaviour in various traffic/situation environments based on their prioritisation between the above two perceptual entities. Implications of the proposed framework at a theoretical and practical level, in view of the future of driving with multiple levels of automation, are also discussed.

Ergonomics for Impartiality and Efficiency in the Law-Courts of Ancient Athens

The judicial procedure in the classical Athenian time (508–323 B.C.E.), was designed to achieve societal values such as representativeness, legitimacy and impartiality. In the present paper, the procedure for selecting and allocating daily jurors to law-courts and the artefacts involved are described, based on the description made by Aristotle in his book Constitution of Athens, as well as findings from excavations in ancient Agora of Athens. It is demonstrated that the whole system meets contemporary ergonomic principles for user-centred and service design. Furthermore, it is shown that the system of the ancient Athenian law-courts represents an exemplar case of a highly sophisticated sociotechnical system, designed based on the understanding of societal values and the users’ micro-moments throughout their entire journey.

Naturalistic Observation of Interactions Between Car Drivers and Pedestrians in High Density Urban Settings

Interactions among drivers and pedestrians especially in heavy urban traffic constitute a key issue that needs to be addressed in the future autonomous vehicles. There is little evidence, however, concerning the signals and cues used by the drivers to infer the future intention of a pedestrian and/or a pedestrian’s awareness of the driver’s vehicle. The paper reports preliminary findings of an instrumented observational study of naturally occurring vehicle – pedestrian interaction cases at high density un-signalized urban crossings. Specifically, 21 experienced drivers drove their own car in a predefined course while equipped with an eye-tracker. In total 321driver – pedestrian interaction cases were analysed based on driver’s eye-gaze analysis and video-assisted retrospective commentary. Several types of signals were identified. These were stratified according to their expressiveness/explicitness. A main finding is that cues with a medium level of expressiveness/explicitness (i.e. eye-gaze from pedestrians) seem to resolve a great number of interaction cases, therefore it is important to explicitly consider this type of cues in future autonomous vehicles. The paper ends with a working model depicting the possible states of mutual attentiveness between driver and pedestrian as identified from the observation data.

Adaptive Warning Strategies from Multiple Systems: A Simulator Study with Drivers with Different Reaction Times

Adaptive interfaces are being developed to avoid drivers’ overload and distraction. 24 drivers, assigned in two groups according to their braking reaction time, participated in a driving simulator study experiencing incidents of concurrent warnings by two support systems. Warnings were provided either independently or via an adaptive interface in which one audio warning was intensified and the other was suppressed. The driving behaviour of the two groups was different, drivers with longer reaction times should be specifically considered when designing adaptive interfaces. The employed adaptive strategy caused changes in the driving behaviour of participants with shorter reaction time, another adaptive strategy, possibly generating warnings earlier, may be more appropriate for drivers with longer reaction times. The metrics that were more sensitive in identifying changes in driving behaviour are mean speed during incident, standard deviation of speed, standard deviation of lateral position and minimum time headway to lead vehicle.

On-road pilot study on the need for integrated interfaces of in-vehicle driver support systems

An on-road experiment has been performed with an equipped vehicle, to study whether the effects on driving behaviour and acceptance of a forward collision warning system and of a lane deviation warning system are different when the systems are isolated or when they are used in parallel. 24 participants were assigned in three experimental and one reference group and were asked to drive the equipped vehicle for 15 consecutive trips on a highway at similar traffic and environmental conditions. The effects of the two isolated systems improve the longitudinal and lateral driving behaviour respectively and are rated as useful and satisfactory, while the use of the systems in parallel does not have a positive effect on driving behaviour. In the latter case the systems are not considered satisfactory and cause frustration to the drivers, thus the need emerges to integrate systems and interfaces.

Understanding overtaking, beyond limitations of the visual system in making spatiotemporal estimations

Motivation -- To enrich our understanding of the factors that influence the decision to overtake against two way traffic, and this, beyond the limitations of the human visual system in making spatiotemporal estimations. Such understanding may be valuable for the design of future adaptive aid systems. Research approach -- An explorative naturalistic observation was conducted with a vehicle equipped with three cameras travelling at normal speed for a total distance of 300 km. 40 unobtrusive observations of overtaking episodes were recorded. Findings/Design -- The data is still at the analysis stage. There is however evidence that before the decision to overtake (i) there is a preparation phase prior to the initiation of the overtaking manoeuvre which deals with the intentions/state of the driver ahead and (ii) frequent users of the particular road will use their specific knowledge and initiate overtaking manoeuvres Research limitations/Implications -- The study has an explorative character mainly for generating hypotheses and cannot as such prove its findings without subsequent complementary methods. Originality/Value -- A contribution of the present paper is on the observation method which ensures the collection of data with a high degree of ecological validity. Take away message -- Spatiotemporal estimation just prior to manoeuvre initiation is only one of the factors influencing the decision to overtake. For understanding naturally occurring overtaking manoeuvres longer time frames of analysis are needed