3D Bicycle-Model Path Tracking
A 3D autonomous vehicle steered by a real Ackermann bicycle model — wheelbase, clamped steering angle, rate-limited actuator — drives itself with a pure-pursuit controller, sensing obstacles only inside a forward LIDAR cone of tunable range and field of view.
A 3D-rendered autonomous vehicle drives itself around a field of pillar obstacles toward a sequence of goal beacons, steered by a real kinematic bicycle model — fixed wheelbase, clamped steering angle, a rate-limited steering actuator — not a free-floating object that turns on a dime. It senses the world only through a forward LIDAR cone of tunable range and field of view, visualised as a rotating ray fan on the ground plane; a potential field turns what the sensor currently reports into a local target, and a pure-pursuit controller converts that target into the steering angle the vehicle's real turning geometry can achieve. Orbit the scene freely while the controller drives.
Simulate bicycle-model vehicle kinematics with clamped steering and pure-pursuit path tracking, avoiding obstacles via a forward LIDAR sensor cone.
3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install