Understanding the role of partially automated systems in improving older driver safety

70岁及以上的成年人在交通路口撞车的风险极大。年长的司机已被证明在十字路口执行适当的扫描行为有困难。已经提出了几种解释，包括上半身活动能力、工作记忆能力、注意相关刺激的能力以及同时处理中央和外围信息的能力。随着技术的迅速发展，几个自动化系统可以通过简化驾驶任务及其要求来帮助人类表现。然而，考虑到老龄化人口倾向于处理的各种问题，包括认知、视觉和运动衰退，显然需要一种自适应的环境界面，该环境界面需要用户付出最少的努力，但更重要的是，可以通过适当的定时警告和警报有效地让用户参与到危急情况中。拟议中的项目不仅寻求了解自动驾驶系统的可用性是否可以提高老年司机在十字路口的安全性，更具体地说，还寻求开发和评估针对老年司机个性化的碰撞警告系统原型。该计划将1)为老年司机开发环境智能碰撞警告界面，以提高十字路口和其他碰撞情况下的安全性；2)开发、测试、集成和验证预测和自适应学习的有效算法；3)反复评估警告系统原型对老年司机安全的有效性，通过眼动、车辆轮廓、生理数据和主观测量来衡量。研究目标将通过四项任务实现，这四项任务将涉及高素质人员接触几个领域，包括但不限于实验设计、机器学习、嵌入式系统、人为因素、认知心理学、仪器和计算机视觉。拟议的项目将应用定性和定量的实验技术，涉及使用多个传感器进行数据收集，收集人类参与者的实验数据，并解决数据同步和解释方面的挑战。这项研究将通过增加老龄化人口和其他司机的安全，对公众产生积极的健康影响。由此产生的知识和技术将代表着在老龄化、人为因素、安全和自动驾驶汽车的交叉点上的重要进步。这项拟议的研究将提升加拿大在地面交通人为因素方面的领先地位，并将使全球驾驶社区受益。

Adults 70 years old and older are at a greatly inflated risk of crashing cars at traffic intersections. Older drivers have been shown to have difficulty with executing appropriate scanning behaviour at intersections. Several explanations have been proposed, including decreases in upper torso mobility, working memory capacity, the ability to attend to relevant stimuli, and the ability to process information simultaneously, both centrally and peripherally. With the rapid proliferation of technology, several automated systems can aid human performance by simplifying the driving task and its requirements. However, given the vast variety of issues that the aging population tends to deal with, including cognitive, visual, and motor declines, there is an evident need for an adaptive ambient interface that requires minimal effort on the part of the user but more importantly can effectively engage the user in critical situations through appropriately timed warnings and alerts. The proposed program not only seeks to understand whether the availability of autonomous systems can improve older driver safety at intersections, but more specifically seeks to develop and evaluate a prototype collision warning system that is personalized to older drivers. The program will 1) develop an ambient intelligent collision warning interface for aging drivers to improve safety at intersections and other crash situations; 2) develop, test, integrate, and validate efficient algorithms for predictive and adaptive learning; and 3) iteratively evaluate the effectiveness of the warning system prototypes on older driver safety as measured by eye movements, vehicle profiles, physiological data, and subjective measures. The research objectives will be accomplished through four tasks that will involve the exposure of highly qualified personnel to several domains including, but not limited to experimental design, machine learning, embedded systems, human factors, cognitive psychology, instrumentation, and computer vision. The proposed project will apply qualitative and quantitative experimental techniques, involve the use of multiple sensors for data collection, collect experimental data from human participants and address challenges in data synchronization and interpretation. The research will yield a positive health impact on the public by increasing the safety of aging populations and other drivers. The resultant knowledge and technology will represent important advancements at the intersection of aging, human factors, safety, and autonomous vehicles. The proposed research will advance Canada's position as a leader in surface transportation human factors and will benefit the driving community globally.