Are own-vehicle images necessary for E-mirrors? Effects of own-vehicle images on driving performance, safety, gazing behavior, and preference

Abstract

Automotive E-mirrors, camera-based mirrorless systems, have started to substitute for conventional outside rearview mirrors. This study examined whether it is desirable to include own-vehicle images in E-mirror screen images for driving performance (lane change time), safety (mental workload assessed using NASA-TLX and occurrence of car accidents), gazing behavior (E-mirror dwell times [DTs] and ratios), and driver preference. A fixed-base driving simulator was used for this study, and two 8-inch tablet PCs were used as E-mirror displays. A total of 27 young drivers made left and right lane changes under four different E-mirror configurations: (1) a control condition (conventional side mirror location and original own-vehicle images; LCIO), (2) LAIO (alternative location and original own-vehicle images), (3) LAIA (alternative location and alternative own-vehicle images), and (4) LAIN (alternative location and no own-vehicle images, where the alternative E-mirror locations were the dashboard top areas next to each side of the steering wheel and the alternative own-vehicle images were the outlined partial own-vehicle images filled in black). Although not significant, LAIN had the shortest mean lane change time. The effect of E-mirror configuration × lane change direction on the mental workload was significant, with the lowest mean mental workload for “LAIN × right lane change”. Furthermore, LAIN involved no car accident, whereas LAIO and LAIA each involved one accident during a left lane change. Although not significant, LAIN had the shortest mean DT. Finally, the effect of E-mirror configuration was significant for driver preference. LAIO and LAIN were the most and second-most preferred configurations, whereas LAIA was the least preferred. Thus, installing E-mirrors at the alternative locations and excluding own-vehicle images (LAIN) is recommended considering driving performance, safety, gazing behavior, and driver preference.