수동 및 자율주행 환경에서 다중채널 경보의 효과 평가

Abstract

Objective: This study aims to investigate the effectiveness of multichannel warning in different levels of automated driving environment. Background: As the driving environment changes from manual to fully automated setting, maintaining the driver's situation awareness and notifying critical information to the driver gets more important. Since automated driving enables driver to perform other tasks than driving, it is important to give a timely notification according to the importance of information. Previous studies evaluated multimodal warning signal to reduce the response time and facilitate driver's decision making, but the efficacy of multichannel warning in an automated driving was not fully validated yet. Method: We measured the reaction time to three modalities (visual, auditory and haptic) and their combinations in three driving conditions (manual driving, monitored driving and automated driving) with 32 participants. Results: Statistical analysis result showed that visual channel is much more vulnerable than haptic and auditory channel. Driver's response time is longer in visual channel than auditory and haptic channel at all three driving conditions. Among three driving conditions, manual driving showed less change in response time than other environments. There was an interaction effect between modality and driving conditions. While the difference in modalities was significant at monitored and automated driving, there was no significant difference at manual driving. Participants were most satisfied with auditory stimuli, followed by haptic and visual stimuli. Conclusion: Contrary to our expectation that haptic and visual signals are more effective than auditory signal, in manual driving, drivers responded to auditory signal faster than others. Haptic and auditory signal showed constantly smaller variances than visual signal in all driving environments, which means that drivers can respond to auditory and haptic signal more reliably for all types of driving tasks. Application: The results of this study can help designing an effective warning signal system for automated driving environment.