Project Description:

Across global markets, transportation systems are rapidly evolving toward automation, pervasive sensing, and intelligent decision-making capabilities. These advancements are often designed primarily around traditional metrics, such as safety, throughput, and cost. Modern autonomous and semi-autonomous systems introduce new types of human exposures (e.g., fatigues, cognitive stress, motion discomfort) and new system constraints (e.g., battery degradation, vibration-induced wear, thermal loads). If left unmanaged, these exposures degrade long-term system performance, reduce user trust and adoption, and impose hidden lifecycle and health costs. This project proposes a new research paradigm for Health-Aware and Durable Transportation Systems, enabling through advanced technologies in autonomous driving, edge computing, and optimized machine learning. We envision that transportation systems can be engineered to actively sense, model, and mitigate human and mechanical exposures, turning transportation into a joint human-machine health ecosystem. The research objectives include: 1) develop joint occupant/vehicle exposure models that quantify health and mechanical burdens, 2) enable adaptive autonomy strategies that mitigate cognitive stress, fatigue, and mechanical wear, and 3) build edge computing framework for efficient inference and control. The overview of the proposed project is shown in the figure below. 

Outputs:

The proposed research will produce new scientific knowledge, computational tools, and operational processes for building health-aware and durability-conscious transportation systems. Expected outputs include: 1) new models that quantify transportation-induced cognitive fatigue, motion discomfort, and stress as functions of autonomous driving behavior and trip context; 2) integrated durability models characterizing mechanical and thermal degradation of vehicles and infrastructure subject to real-world roadway exposures; 3) real-time control and planning strategies to optimize for human comfort, safety, and system durability; and 4) edge-optimized machine learning architectures for in-vehicle inference that enable low-latency assessment of human sate and vehicle wear.

Outcomes/Impacts: 

The proposed research is expected to influence both the operational practices of future autonomous vehicle systems and the broader transportation policy ecosystem. Health- and durability-aware sensing and control technologies could be incorporated into commercial autonomous driving stacks and fleet management software, enabling safer and more comfortable travel for passengers, as well as reduced wear and operating cost for fleets. New exposure models for cognitive fatigue, motion discomfort, and vehicle degradation may inform guidelines for ergonomic route planning, autonomous driving style optimization, and proactive maintenance scheduling. These capabilities could motivate new standards for human-centered performance metrics in automated mobility, complementing existing safety benchmarks and contributing to USDOT’s efforts in defining assessment frameworks for autonomous vehicles. In the longer term, validated research results may influence policy and regulatory considerations related to safety, fleet longevity, and deployment of advanced AI and edge computing technologies, ensuring more resilient and human-centered mobility systems.