Risk-Aware Autonomous Driving
This project thread focuses on local planning and decision-making methods for autonomous driving under uncertainty. The core objective is to produce motion plans that remain computationally tractable while respecting explicit risk constraints in dynamic traffic environments.
Core Questions
- How can local planners remain safe under uncertain predictions of surrounding traffic agents?
- How can spatiotemporal risk be represented in a way that supports fast online optimization?
- What theoretical guarantees can be provided for practical real-time planning systems?
Methods
- Graph-based and lattice-based local planning
- Constrained shortest-path formulations
- Competitive and regret-aware online optimization
- Risk-bounded decision-making under prediction uncertainty
Representative Outputs
- Graph-Based Local Planning with Spatiotemporal Risk Assessment for Risk-Bounded and Prediction-Aware Autonomous Driving
- Online Risk-Bounded Graph-Based Local Planning for Autonomous Driving with Theoretical Guarantees
Research Stack
- CARLA
- ROS 2 and NAV2
- CommonRoad
- Python and C++
- Optimization and simulation workflows