How this sensor-fusion model works

This 2D companion strips the 3D self-driving car model down to the mechanic that actually drives autonomy: a car continuously casts a ring of LiDAR rays, sweeps a camera field of view, and polls a wider-range radar for moving traffic, then fuses all three readings into one live confidence score. Adjust AI Confidence base and Sensor Range and watch the fused confidence, braking behaviour and steering react in real time — exactly the trade-off real autonomous-vehicle stacks have to make between sensor trust and driving caution.

In Autonomous mode the car follows a loop track using pure-pursuit steering toward the next waypoint. When LiDAR detects an obstacle inside the braking zone ahead, the controller blends a steering-away vector with speed reduction, scaled down further when fused confidence is low (a cautious AI slows earlier). Switch to Manual to drive with the arrow keys and see the same sensor dashboard update around your own driving.