危害
危害分析
适应(眼睛)
心理学
计算机科学
毒物控制
社会心理学
环境科学
地理
数据收集
作者
Xu Wang,Lie Guo,Linli Xu,Pingshu Ge,L Guan,Renxiang Yuan
标识
DOI:10.1080/10447318.2026.2625957
摘要
Potential hazard warnings (PHWs) can enhance driving safety in complex scenarios; however, behavioral and safety implications of repeated PHWs remain unclear. This study developed a human-automation interaction model for PHWs in L3 ADS to examine the cry wolf effect and behavioral adaptation, while considering the moderating role of hazard type. We developed a PHW system, conducted a driving-simulator experiment with 92 participants, and validated our model using multilevel structural equation modeling. Results showed that repeated PHWs triggered the cry wolf effect, which degraded attention, vigilance, and situation awareness, thereby undermining driving safety. Moreover, PHWs fostered trust in the system. However, increased trust induced driver behavioral adaptation, reflected in reduced attention and situation awareness. Such adaptation imposed an additional safety impairment equivalent to 48.8% of that caused by the cry wolf effect. Furthermore, multi-group analysis revealed how hazard type moderated human-automation interaction. These findings inform safe deployment of PHWs in ADS.
科研通智能强力驱动
Strongly Powered by AbleSci AI