Modeling of Human Balance and Step Recovery in Public Transportation

Research Project, 2027 – 2028

Free-standing passengers on public transportation experience various unexpected disturbances caused by road conditions and driver maneuvers during their commute, which might be difficult to stabilize, especially for passengers with limited muscle strength, and thus increasing the risk of falling into surrounding structures or other passengers which can lead to injury. The safety evaluation in such cases is highly challenging due to lack of simulation models (generalizable over demographics) and challenges in collecting realistic and extensive volunteer data. Recent research in legged robotics has shown that it is possible to obtain very natural sensorimotor control in humanoid robots. These systems can perform robust balance, step recovery, walking and running tasks. This is made possible by significant advancements in reinforcement learning, real-time whole body model predictive control (MPC), computationally efficient differentiable multi-body dynamics and the evolving capabilities of the computing hardware. This project proposes the transfer of such methods in simulating the closed loop dynamics of human balance and step recovery inside urban public transportation systems such as city trams, buses etc. The project aims to deliver a parameterized human model which can be used to simulate humans in different age groups with varying mass distributions and muscle strengths. The benefits of this project are threefold. First, the project deliverables will help better design and evaluate autonomous and driver support functions in public transport vehicles. Second, these models can be used to improve planning of roads and tram lines and generate better guidelines for path-dependent speed limits for such vehicles. Lastly, the results of this project will also be beneficial for the robotics community as this can help them design better rehabilitation aids (for e.g., exoskeletons) for improving the mobility of elderly and disabled passengers. The work combines methods from transport safety, robotics, biomechanics, and human factors, and directly contributes to the AoA Transport vision of safe, efficient, and inclusive transport systems.