A hands-on workshop for VISST School robotics students
Learn how robots combine multiple sensors to know where they are!
Every sensor lies. No sensor is perfect:
| Sensor | Strength | Weakness |
|---|---|---|
| IMU / Gyro | Smooth, fast | Drifts over time |
| Wheel Encoders | Always available | Wheel slip |
| Camera / AprilTags | Absolute position | Noisy |
Combine sensors with opposite weaknesses:
| IMU / Odometry | Camera / AprilTags | |
|---|---|---|
| Noise | Low | High |
| Drift | High | None |
| Combined | Low | None |
PREDICT (odometry) ──► UPDATE (camera) ──► PREDICT ──► ...
uncertainty ↑ uncertainty ↓
| Time | Topic |
|---|---|
| 10 min | Why Sensor Fusion? |
| 15 min | The Kalman Filter |
| 25 min | Hands-on Demos |
| 10 min | FRC Integration |
Workshop by Grigory Artazyan & Claude