Self-directed project
Machine Learning for Automotive Safety
Perception Validation Engineer (ADAS Safety Sign-off) · AI / ML · August 2026
A ProoV case study · educational project, not employment
Built a perception validation dossier for a synthetic Aurora-7 detector with correct IoU-based detection rules, honest precision/recall scoring, and a safety-focused VRU false-negative analysis. Sliced recall by lighting, visibility, and rarity to surface the hardest pedestrian and cyclist cases, then mapped the real Uber ATG Tempe crash to the same night-time pedestrian failure mode using NTSB evidence.
Graded against
- Detection metrics correctness (precision, recall, VRU false-negative rate)30%
- Failure-mode analysis & per-condition rigour25%
- SOTIF safety case quality (ISO 21448 reasoning)25%
- Ship verdict — evidence-based & internally consistent20%
Passed · pass mark 60/100
What stood out6
- Correctly elevated VRU false-negative rate as the safety-critical metric instead of headline accuracy
- Exposed the night-time and occluded VRU collapse with concrete recall numbers, including the worst night+occluded slice at 0.27
- Anchored the Tempe crash to the night-time pedestrian miss row and the NTSB timing evidence
- Kept the detection definition consistent with IoU >= 0.5 and unmatched real objects as false negatives
- Leakage-Free Safety Metric Framing
- Condition-Sliced Failure Analysis
The work I submitted18 tasks
10 tasks · 11.8k characters · Python · 8 written answers · 808 words
- Sensor Match553 characters
- Example Iou Demo1.5k characters
- Task Definition378 characters
- Example Metrics DemoPython · 2.2k characters
- Scored Metrics RunPython · 2.3k characters
- Scored Metrics351 characters
- Ttcs Example SlicePython · 1.3k characters
- Ttcs Triage SlicePython · 1.6k characters
- Realcase Analysis848 characters
- Teachback831 characters
- Sensor Match27 words
- Example Iou Demo169 words
- Task Definition42 words
- Example Metrics Demo154 words
- Scored Metrics Run143 words
- Scored Metrics18 words
- Ttcs Example Slice143 words
- Ttcs Triage Slice112 words
Summarised on purpose — the submitted code and writing stay private so this page cannot be reused as an answer key. The full submission sits behind the verified certificate.