Visual Attention of Young and Older Drivers in Takeover Tasks of Highly Automated Driving

Abstract

The objective of this study is to examine the differences of visual attention patterns between young and elder drivers affected by engagement of audio non-driving related task (NDRT) and traffic conditions. Previous research focus on influence of the interaction of NDRT in Highly Automated Vehicle (HAV) during take over tasks both in driving performance and visual attention, yet the research on aging effect on visual attention pattern with NDRT during take over tasks is still limited. Twenty-seven young and twenty-seven elder participants drove in a highly automated driving simulator in different traffic conditions and NDRT levels, and after the take-over request (TOR) they were required to change lanes to pass a stationary truck. The traffic conditions are set by numbers of cars on the right lane and NDRT is classic n-back task with audio stimuli. Results showed that elder drivers have higher collision rate and less stable lateral control than young drivers. For fixation transition number between different regions, both age group have fewer transitions when engaging NDRT, elder group have significantly fewer transitions than younger group and are more influenced by high traffic density. The analysis for fixation durations show similar results. For the change of distribution of fixation time, elder group focus more on the road ahead than young group which is more obvious in high-traffic condition. Around 4 s after the TOR, a decrease of attention on mirrors is observed in elder group. These results show a probable difference in visual attention patterns for young and elder drivers when facing TOR with engagement of NDRT and in high traffic density. Possible relevance between less attention on mirrors and worse driving behavior in elder group is also discussed. Further improvements of this study is illustrated according to the previous results and analysis.