Driver cognitive engineering and challenges for the future

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. Still, human behaviour is fairly contextual and substantially different from one person to another. Thus, the initial linear models have been gradually replaced by non-linear and even probabilistic models, based upon AI principles, such as Artificial Neural Networks or Genetic Algorithms. This becomes even more intrigued if we consider such a complex behavioural task, as vehicle driving. The traffic system as a whole can be seen as being comprised of three interactive parts: vehicles, road users and the road environment. Any traffic situation is the result of the interaction between these three systems. Normally the traffic situation develops as planned, but, in certain circumstances the resulting interaction will result in a critical situation or in a crash. The driver is a critical component of the traffic system. Attempts have been made to estimate the importance of the driver as an accident cause. It has been estimated that road user factors are sole or contributory factors in a great majority of road crashes. There is no generally accepted model of the complete driving task. There are detailed descriptions focusing on perception and handling aspects, and reporting what drivers really do in every possible (normal) situation from the beginning to the end of a journey. There are also more analytical approaches focusing on driver behaviour in relation to task demands, with the purpose of trying to explain and understand the psychological mechanisms underlying human behaviour. Usually, car driving is described as a task containing three different levels of demands. At the strategic level, the general planning of a journey is handled. For example, the driver chooses the route and transportation mode, and evaluates resulting costs and time consumption. At the tactical level, the driver has to exercise manoeuvres, allowing him/her to negotiate the right now prevailing circumstances, for instance turning at an intersection or accepting a gap. Finally, at the control (stabilisation) level the driver has to execute simple (automatic) action patterns, that together form a manoeuvre. Examples are changing the gear and turning the wheel. The demands imposed on the driver are met through his/her driving behaviour. Also, the performance of the driving task is usually assigned to three different levels: knowledge-based, rule-based and skill-based behaviour. Skill-based behaviour is described as data-driven, meaning that skills are performed without conscious control and use of attention resources. They are immediate and efficient. Rule-based behaviour on the other hand occurs under conscious control and requires attention. Therefore, it is less immediate and efficient. Knowledge-based behaviour involves problem solving and is relevant when it is not given how to act in a specific situation. Thus, an important aspect of knowledge-based behaviour is that reasoning is required (A). For the covering abstract of the conference see ITRD E212343.

  • Authors:
    • GIANNOPOULOS, G
    • PANOU, M
    • BEKIARIS, E
  • Publication Date: 2006

Language

  • English

Media Info

Subject/Index Terms

Filing Info

  • Accession Number: 01102533
  • Record Type: Publication
  • Source Agency: TRL
  • Files: ITRD
  • Created Date: Jun 16 2008 7:54AM