y0news
← Feed
Back to feed
🧠 AI NeutralImportance 7/10

CRASH: Cognitive Reasoning Agent for Safety Hazards in Autonomous Driving

arXiv – CS AI|Erick Silva, Rehana Yasmin, Ali Shoker|
🤖AI Summary

Researchers introduced CRASH, an LLM-based agent that analyzes autonomous vehicle incidents from NHTSA data covering 2,168 cases and 80+ million miles driven between 2021-2025. The system achieved 86% accuracy in fault attribution and found that 64% of incidents stem from perception or planning failures, with rear-end collisions comprising 50% of all reported incidents.

Key Takeaways
  • CRASH AI agent achieved 86% accuracy in analyzing autonomous vehicle crash reports from expert validation.
  • 64% of AV incidents attributed to perception or planning system failures, highlighting critical areas for improvement.
  • Rear-end collisions represent 50% of all reported AV incidents, indicating a persistent challenge in autonomous driving.
  • The system processed over 2,168 real-world cases representing more than 80 million autonomous miles driven.
  • LLM-based analysis enables scalable and standardized investigation of AV incidents across different manufacturers.
Read Original →via arXiv – CS AI
Act on this with AI
Stay ahead of the market.
Connect your wallet to an AI agent. It reads balances, proposes swaps and bridges across 15 chains — you keep full control of your keys.
Connect Wallet to AI →How it works
Related Articles