If Perception is the eyes of the car, Planning & Control is the brain and hands—deciding the path and steering the vehicle safely through the track.
Trajectory Generation
The Planning & Control (P&C) team transforms cone detections into smooth, drivable trajectories. By fusing vehicle motion data with cone positions, we filter noise, estimate the track centreline, and generate cubic-spline trajectories. These are followed using a pure-pursuit controller that converts the path into safe, stable steering commands.
Focus Areas
Trajectory Generation: Using cubic spline interpolation for smooth centreline paths.
Control Algorithms: Implementing pure-pursuit with PD feedback and low-pass filtering.
Noise Robustness: Evaluating steering accuracy under Gaussian cone detection noise.
Practices
Reliability First: Designing controllers that remain stable under perception noise.
Safety Constraints: Bounding steering commands to ensure feasible vehicle dynamics.
Close Integration: Working hand-in-hand with Perception for real-time data fusion.
Iterative Testing: Validating algorithms in simulation before on-track deployment.
Typical Deliverables
Trajectory Planner: Centreline paths generated from cone detections.
Controller Modules: Pure-pursuit and PD steering controllers with filtering.
Simulation Reports: RMSE and robustness evaluations under varying noise.