• DocumentCode
    529142
  • Title

    Toward inference of driver´s lane-change intent under cognitive distraction

  • Author

    Zhou, Huiping ; Itoh, Makoto ; Inagaki, Toshiyuki

  • Author_Institution
    Dept. of Risk Eng., Univ. of Tsukuba, Tsukuba, Japan
  • fYear
    2010
  • fDate
    18-21 Aug. 2010
  • Firstpage
    916
  • Lastpage
    920
  • Abstract
    This paper investigated driver glances in order to establish a method for inferring lane-change intent when a driver was performing a cognitively distracting task. A driving simulator experiment was conducted for collecting driver glances´ data under a single-task (changing lanes) condition as well as a dual-task (the changing lanes and performing a secondary cognitive task) condition. The degraded detecting ability was shown on the intent inference under the dual-task condition. Comparisons of the glances´ data between the two driving conditions revealed that the insufficient glances of looking at the side-view mirror and the delayed lane-change intent might cause the decrease of the detection rate. To establish an appropriate method to infer the lane-change intent for supporting a driver under such distracted conditions, we investigated the lane-change intent under different conditions, and the result showed the necessity of adjusting the inference algorithm for responding a driver who might be distracted by a non-driving cognitive task. This finding indicated that it was important to infer the correct driver intent for establish a support function system for avoiding a lane-change crash due to driver distraction.
  • Keywords
    cognition; driver information systems; inference mechanisms; cognitive distraction; driver lane-change intent; driving simulator experiment; inference algorithm; nondriving cognitive task; Analysis of variance; Delay; Driver circuits; Lead; Mirrors; Vehicle crash testing; Vehicles; driver glances; driver intent; lane changes; safety system;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    SICE Annual Conference 2010, Proceedings of
  • Conference_Location
    Taipei
  • Print_ISBN
    978-1-4244-7642-8
  • Type

    conf

  • Filename
    5602300