Design of an Experimentation Platform to study Take-over in Human Driver-Automated System Transition

Dogan D., ACARMAN T., Estrada O. S.

31st IEEE International Symposium on Industrial Electronics, ISIE 2022, Alaska, United States Of America, 1 - 03 June 2022, vol.2022-June, pp.182-187 identifier identifier

  • Publication Type: Conference Paper / Full Text
  • Volume: 2022-June
  • Doi Number: 10.1109/isie51582.2022.9831632
  • City: Alaska
  • Country: United States Of America
  • Page Numbers: pp.182-187
  • Keywords: human factors in vehicular system, take-over request time, sensors, and electric vehicle
  • Istanbul Technical University Affiliated: Yes


© 2022 IEEE.This paper contributes to the experimental take-over request (TOR) study. For emerging autonomous driving applications, determination of TOR is time critical when control authority transition from the automated driving system to the human driver in conditional automation is occurred. For each driver, TOR time is personalized by analyzing the TOR transient responses during the driving experiments and answers to the surveys. The automated system is a prototype electric vehicle with longitudinal autonomy. A set of sensors constituted by an encoder, an inertial measurement unit (IMU), a current measurement sensor, a global positioning system (GPS) receiver and a galvanic skin response (GSR) sensor are used for the vehicle data collection system for the study. A human-machine interface (HMI) is designed on a smart phone to transmit messages about driving tasks to be accomplished by the driver to enrich the TOR study. Finally, the above designed TOR experimental study is applied as a case study for 3 different driver groups (experienced, semi-experienced and inexperienced drivers) in 4 different TOR times (Os, 2s, 4s, 6s). The variances of ideal TOR times of experienced, semi-experienced and inexperienced drivers in the experimental campaign are calculated as 0.94, 2.05 and 2.42, respectively.