Robust control and safety assurance for over-actuated autonomous electric vehicles

Research Project, 2026 – 2027

This project has three goals: 1) develop a modular and scalable control framework for over-actuated electric vehicles (EVs), ensuring real-time implementability and adaptability across different vehicle platforms; 2) design predictive, transparent, and verifiable trajectory planning and failsafe control algorithms for autonomous motion planning and force allocation of over-actuated EVs, ensuring robustness to noise, disturbances, and system failures; and 3) establish a predictive safety evaluation method using counterfactual simulations, fault injection and real-world crash data to assess accident prevention and failsafe operation of over-actuated EVs. These goals will solve the problem of safer autonomous motion planning and force allocation of emerging over-actuated EVs that are posed to implement new vehicle technologies, including torque vectoring, electromechanical brake-by-wire and steer-by-wire. They will also solve one of the key challenges in the event of a steering actuator failure, where the vehicle must remain safe and controllable despite the loss of direct steering input.