Driver impairment states in real traffic: identification and countermeasures - the new safety challenge in vehicle-human reliability (résumé)
KHARDI ; VALLET
Type de document
COMMUNICATION AVEC ACTES INTERNATIONAL (ACTI)
KHARDI ; VALLET
Résumé / Abstract
Drivers impairment on the wheel is a frequently reported phenomenon in car driving. They are suspected to play an important role in occurrence of fatal driving accidents with some non inconsiderable percentage. The drivers impairment monitoring systems promise safety through computer-based decision and automation. However, there is evidence that people are not good monitors of automation. Perhaps drivers become inattentive to the automation and are easily distracted to tasks in which they are obliged to play a more active role, or they become bored and fall asleep if they are not demanding tasks. This phenomenon is well established from observation. For this we pay great attention to system reliability which analysis is expensive and tedious world the question is, will the availability of drowsy-drivers detection systems and other ivhs safety advantages leads drivers to take additional risks and compromise the gains in safety ? The aim of our research is to use new technologies and to develop efficacious methods in physiological and car mechanical settings treatment that have the potential to increase safety and reliability of systems in the short term. The present study characterizes vigilance at the beginning of drivers drowsiness. Whenever we have summoned vigilance levels, they are certainly synonymous with driver behavioural impairment states. We have examined the state of the art in monitoring driver impairment with a view to establish the feasibility of carrying out a vigilance states detection device. We have reviewed and discussed empirical studies that assessed the relationship between physiological indices, measurements of car mechanical parameters which are used as a criteria in driver vigilance states assessment and driver' actions. Our paper permits to compare several parameters in real traffic. Some of them would be able to be enclosed in vigilance levels detection subsystem and are assumed to be sensitive in the driver behavioural state changes. We affirm the role of some indices in the precise identification of drivers' impairment linked with drowsiness.