Lab 10 — Final Autonomous Fleet Challenge¶
The challenge¶
Full mission, multiple cars, shared arena:
Every subsystem from Weeks 1–9, running together, unattended.
What is assessed¶
TODO — instructor: set the rubric and the pass bar. A workable split:
| Component | Weight | What it means |
|---|---|---|
| Mission completion | TODO % | Pickup to release, unattended |
| Safety | TODO % | No contact between cars, no contact with obstacles |
| Telemetry | TODO % | Dashboard reflects true vehicle state throughout |
| Repeatability | TODO % | Three consecutive runs, not one |
| Report & analysis | TODO % | Performance data, and honest failure analysis |
Require three consecutive successful runs
A robot that works once got lucky. Demanding three in a row is what forces teams to fix the intermittent problems they have been working around all semester — and the gap between one success and three is where most of the engineering learning in this course actually happens.
Performance results to report¶
TODO — specify. Suggested measurements:
- Mission completion time, per run.
- Position error at pickup and at delivery.
- Number of obstacle stops, and number of coordination conflicts resolved.
- Any run that failed, and the diagnosis.
Failed runs are data
A report that says "run 2 failed because the heading estimate drifted 15° after the third turn, which we traced to gyro bias re-emerging as the board warmed up" is worth more than a report with three clean runs and no analysis. Say so in the rubric, or students will hide their failures.
Deliverable¶
Final fleet demonstration + performance results.