A Hybrid Target Selection Model of Functional Safety Compliance for Autonomous Driving System

What is it about?

Self-driving cars struggle to quickly identify key obstacles (e.g., pedestrians or sudden stops) in heavy traffic. Existing methods either use rigid safety rules (which fail in complex situations) or opaque AI models (which can’t explain decisions). Our solution combines both approaches like a driver using GPS and road signs together: simple rules handle clear scenarios, while a streamlined AI learns from past driving patterns. This hybrid system makes safer decisions in cities/highways than either method alone, using minimal computing power. Real-world tests confirm it reliably responds to dangers like sudden braking or lane-cutting cars, proving both effective and practical for everyday use.

Featured Image

Photo by Anouk Fotografeert on Unsplash

Photo by Anouk Fotografeert on Unsplash