原文传递 Distraction Detection and Mitigation Through Driver Feedback. Appendices. Final Report, September 5, 2008-April 4, 2011.
题名: Distraction Detection and Mitigation Through Driver Feedback. Appendices. Final Report, September 5, 2008-April 4, 2011.
作者: Brown, T. L.; Davis, C.; Lee, J. D.; Liang, Y.; Marshall, D.; Moeckli, J.; Nadler, E.; Roberts, S. C.; Schwarz, C.; Victor, T.; Yekhshatyan, L.
关键词: Distracted Driving; Driver Feedback; Driver Monitoring Systems; Evaluation Protocol; Real-time Driver Monitoring; Safety Benefits
摘要: Despite government efforts to regulate distracted driving, distraction-related fatalities and injuries continue to increase. Manufacturers are introducing real-time driver monitoring systems that detect risk from distracted driving and warn drivers; however, little is known about these systems. This report identifies evaluation techniques to characterize and assess these emerging technologies, presents results of their application, develops a framework for estimating systems safety benefits, and provides safety relevant information to guide technology development. A standardized language for describing and differentiating systems was created, and its application revealed key trends in the design landscape. A novel approach to detection that provides prospective indications of safety-critical vehicle state changes is described. Two evaluation protocols were developed and to provide empirical assessments of (1) detection algorithm performance and (2) the effect of mitigations on driver performance and acceptance. The protocol included driving on different types of roadways and performing secondary tasks in the high-fidelity NADS-1 driving simulator. Four progressively complex distraction detection algorithms were compared to evaluate the ability of vehicle-based systems to distinguish between distracted and non-distracted drivers. Algorithm performance varied across road types and distraction tasks. A safety benefits framework appropriate for distraction mitigation systems is proposed that expands on past benefit analyses.
报告类型: 科技报告
检索历史
应用推荐