Due to substantial gains in road safety on many factors, the share of accidents caused by defects of attention increases.
The issue of the ATLAS project is to estimate the on-accident risks associated with attention flaws and understand their causes or origins and their effects on driving.
This project offers an innovative approach to leverage knowledge for the development of driver monitoring system. In this context, only supervised learning techniques used to identify driver distraction (support vector machine and Bayesian network) obtain encouraging results but these analyses are still often too difficult because the effect of inattention can vary greatly from a driving situation to another and from one type to another conductor. By developing an experimental approach to master certain parameters of influence, this project will allow substantial progress.
Epidemiological findings show that careful driving is a traffic safety deposit likely to reduce the number of casualties on our roads. The experimental results show the influence of various types of cognitive control (attention control, emotional regulation and behavioural inhibition). Different control strategies of cognitive effort in driving were described. This project has confirmed a strong public health concern on this theme. Prospects to make the least costly driving attention plan are described and a critical inventory of methods to raise new technological barriers.
The exploratory analysis techniques of sequence data and supervised conduct learning applied to data collected on the motorway will search algorithms capable of identifying a distracted driver, psycho-ergonomic analysis of driving activity to enrich driving data. New technological challenges, constituting the audience to adapt depending on the condition of the driver, and will soon be overcome thanks to the multidisciplinary approach advocated in this project.