Authors: Anuj K Pradhan (University of Massachusetts Amherst) , Jacob Crossman (Soar Technology) , Adam Sypniewski (Deepgram)
Advanced technologies such as adaptive cruise control and lane keeping are key components of SAE Level 2 vehicle automation. As such automation becomes widespread, drivers may be less engaged in driving because they assume that vehicles can safely mitigate risks. However, L2 automation cannot handle the full spectrum of driving situations and will require manual control in many situations. Drivers unprepared to take control may make suboptimal, delayed, or dangerous decisions during and after reengaging with the driving task. This highlights the need for efficient ways to help drivers re-engage with driving. This paper describes an evaluation of a conceptual driver engagement system that combined driver data with contextual data to communicate appropriate information during L2 operations. The system was compared to a traditional, staged-alert system that only monitored driver gaze with no contextual information. Results indicate higher situation awareness, higher levels of trust and satisfaction, no increase in workload, with evidence of improve off-road glance behaviors when driving with the conceptual system. These findings can help inform further development and testing of driver engagement approaches using driver monitoring.
How to Cite: Pradhan, A. , Crossman, J. & Sypniewski, A. (2019) “Improving Driver Engagement During L2 Automation: A Pilot Study”, Driving Assessment Conference. 10(2019). doi: https://doi.org/10.17077/drivingassessment.1707