Evidence Collection: Automotive Driving Aids and Liability

### Forensic Analysis of Automotive Driving Aids At 00:00, automotive and SCADA security researcher Vlad examines the forensic and liability challenges surrounding Advanced Driver Assistance Systems (ADAS) and semi-autonomous automotive platforms.

### Sensor Malfunctions and Control Unit Segfaults At 04:00, Vlad explores failure states in automotive computer systems. He details how malformed CAN bus frames or unhandled sensor edge cases can trigger segmentation faults or freeze driver-assist microcontrollers, preventing safety systems from executing timely steering or braking corrections.

### Deconstructing Event Data Recorders (EDRs) At 08:03, the presentation analyzes the vehicle "black box," formally known as the Event Data Recorder (EDR). Vlad explains the trigger algorithms that initiate recording, data write cycles to non-volatile EEPROM, and the physical protections designed to survive high-impact collisions.

### Overcoming Data Extraction and Integrity Hurdles At 12:03, Vlad highlights forensic extraction pitfalls. Connecting standard diagnostic tools through the OBD-II port can inadvertently alter volatile memory registers or induce state changes. He demonstrates non-invasive chip-off extraction and direct EEPROM probing techniques that preserve chain-of-custody and evidence integrity.

### Telemetry Reconstruction in Collision Inquiries Concluding between 16:00 and 20:08, the speaker demonstrates accident reconstruction methodologies. By synchronizing steering angle logs, throttle position sensors, and autonomous emergency braking (AEB) timestamps, forensic investigators can determine whether a collision resulted from driver negligence or autonomous system failure.

(AI summary — Gemini Flash)