| R01 | Perception System Underperformance | Technical | L2 × C5 → L3 × C4 | - Already have perception pipeline working for SVD so less likely to fail
- Collect diverse data varying lighting and shadow conditions
- Anomaly detection pipeline causes occasional false positives with high processing time.
- Change the core architecture to fit into lab scenarios.
|
| R02 | End-Effector Breakage | Technical; Schedule | L5 × C1 → L4 × C1 | - We have a metal + 3d printed mechanism to make sure it’s easily replaceable when broken
- Likelihood reduced because metal+3d gripper proved robust.
|
| R04 | Hardware Damage | Technical; Schedule | L4 × C5 (no change) | - Reserve budget for spares.
- Use emergency-stop mechanisms and run at reduced speed during testing.
|
| R05 | Budget Overruns | Budget | L3 × C2 (no change) | - Maintain a budget buffer and prioritise essential components.
- Leverage sponsor lab resources when available.
|
| R07 | Integration Issues | Technical; Schedule | L3 × C3 → L4 × C4 | - Unforeseen faults only emerge in end-to-end runs.
- Enforce strict version control (Git).
- Run frequent CI-style integration tests across subsystems.
- Run weekly integrated tests and freeze validated fallbacks prior to FVD.
|
| R09 | Reduced Personnel | Schedule | L4 × C3 (no change) | - Plan critical deadlines with margin for flexibility.
- Hold regular scrums to catch delays early.
|
| R10 | Navigation Error | Technical | New: L2 × C3 | - Base stopping off-position can leave the target out of the arm’s reach.
- Measured all relative base positions from which planning works and ensured the max navigation error is less than this range.
|
| R11 | False Verdict | Technical | New: L3 × C5 | - Testing the pipeline on multiple videos and change the architecture as per false verdicts
|
