CAREER: Adaptive Driver Assistance Systems and Personalized Training for Law Enforcement Officers (ADAPT-LEO)

机动车碰撞是执法人员因公死亡的主要原因，占执法人员致命工伤的近40%。在所有职业中，这一比例也是全国平均水平的2.5倍。这些撞车事故的主要原因包括警官在驾驶时使用车载技术、疲劳以及缺乏足够的培训来处理执法中出现的高要求驾驶情况。该项目将模拟警察在高需求情况下的驾驶工作量和表现，然后利用这些模型开发适应警察工作量的车载技术和培训解决方案，以减少警察行动中与撞车有关的伤害风险。所开发的模型、方法和工具也可能有益于其他驾驶和培训领域。这项工作将支持研究生和本科生的研究培训；此外，项目组将为执法小组和K-12教师编制外联材料，帮助他们利用这项工作支持培训和STEM教育。该项目的技术目标分为三个方面。《第一推力》利用自然主义驾驶研究、知识获取方法和认知建模软件，对新手警察在驾驶时的认知、知觉和运动需求进行建模。推力2号将基于一种混合算法开发和评估自适应车载技术接口，该算法利用计算认知性能模型、机器学习算法以及实时行为和生理测量来提高军官的驾驶性能，减少精神负担和分心。Struts 3将根据军官的认知状态以及实时和离线性能测量来开发基于自适应驾驶模拟的培训，评估基于自适应驾驶模拟培训在改善驾驶、次要任务表现和学习方面是否比之前基于操作驾驶性能或视觉注意力的自适应培训协议更有效。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Motor vehicle crashes are a leading cause of line-of-duty deaths for law enforcement officers, accounting for almost 40% of officers' fatal work injuries. They are also 2.5 times more frequent than the national average among all occupations. The main contributors to these crashes include officers' use of in-vehicle technologies while driving, fatigue, and lack of sufficient training in handling high-demand driving situations that arise in law enforcement. This project will model officers' driving workload and performance in high-demand situations, then use those models to develop in-vehicle technology and training solutions that adapt to officers' workload in order to reduce the risk of crash-related harms in police operations. The models, methods, and tools developed may also benefit other driving and training domains. The work will support the research training of graduate and undergraduate students; further, the project team will develop outreach materials for both law enforcement groups and K-12 teachers to help them use the work to support training and STEM education. The technical aims of the project are divided into three thrusts. Thrust 1 focuses on modeling novice police officers' cognitive, perceptual, and motor demands while driving, using the findings of a naturalistic driving study, knowledge elicitation methods, and cognitive modeling software. Thrust 2 will develop and evaluate adaptive in-vehicle technology interfaces based on a hybrid algorithm that leverages computational cognitive performance models, machine-learning algorithms, and real-time behavioral and physiological measures to improve officers' driving performance and reduce mental workload and distraction. Thrust 3 will develop adaptive driving simulation-based training based on officers' cognitive state and real-time and offline performance measures, assessing whether the adaptive driving simulation-based training is more effective than prior adaptive training protocols based on operational driving performance or visual attention in improving driving, secondary task performance, and learning.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.