Virginia Tech® home

Strategic Safety-Critical Attacks Against an Advanced Driver Assistance System

Paper Details

Abstract

Semi-autonomous vehicles frequently rely on advanced driver assistance systems (ADAS). The increasing complexity and connectivity of ADAS exposes vehicles to attacks. 

Researchers investigated the resilience of a widely used ADAS against safety-critical attacks that target control systems.

 Results highlighted the intolerance of ADAS to safety-critical attacks and the importance of timely interventions by human drivers or automated recovery mechanisms to prevent accidents.

Researchers propose an efficient way to discover the safety vulnerabilities of existing ADAS using a realistic driving simulator to mitigate vulnerabilities before real deployment on the road to save time and reduce the risk of harm in real road testing. 

Results showed the proposed Context-Aware attacks can achieve an 83.4 percent success rate in causing hazards, 99.7 percent of which occur without warning.