PRIVACY IN AUTOMATED AND CONNECTED VEHICLES
summary
Connected-vehicle demos and papers show CAN raw logs reidentify drivers using machine learning and enable micro maneuvers and macro route reconstruction without GPS; insurance firms already deploy insurance dongles collecting OBD/CAN data for pay‑how‑you‑drive policies. Prototypes put biometric templates on a smartphone secure element as a car key and outline a backend message protocol for gated device access. Hardware roots of trust (TPM/enclaves) plus homomorphic encryption are proposed to prove data provenance and compute on encrypted telemetry. New regs cited: UNECE R155 and ISO 21434.
AI-written summary · about
chapters (8)
- 0:00Start
- 8:32How to address these challenges?
- 11:12Automotive data ecosystem (1)
- 14:41Driver re-identification (2)
- 25:33IDEMIA Driver Authentication Demonstrator
- 35:29Driver monitoring to improve safety
- 40:37Trade-off between safety and privacy
- 55:07Summary & Resource List
related talks
01h 13mBuckle-up: connected cars could face privacy bumps in the road ahead.
25mPrivacy Concerns in the Connected Car Ecosystem
36mCar Hacking
37mAutomatic Automotive Hacking
58mA Survey of Remote Automotive Attack Surfaces
38mNation-State at Fault: The Future of Automotive Accidents?
41mPanel: Tackling Automotive Hardware Vulnerabilities
50mAdvanced CAN Injection Techniques for Vehicle Networks