New research reveals critical gaps in autonomous vehicle safety protocols

Researchers Wolfgang Kröger and Leon Johann Brettin from arXiv's Electrical Engineering and Systems Science division have published a critical analysis of the challenges autonomous vehicles (AVs) face when transitioning from Minimal Risk Condition (MRC)—a state where the vehicle can no longer safely operate—to normal driving or a safe stop. The paper, titled *Automated vehicles: Challenging transition from Minimal Risk Condition*, reveals that current regulations and technical systems are inadequate for handling this transition, leaving human operators overwhelmed by complex decisions.

The study identifies nearly 30 distinct failure scenarios where an AV might enter MRC, ranging from sensor malfunctions to software crashes. Human intervention operators must then assess whether to maintain the vehicle’s position (e.g., on a road shoulder), re-enter traffic flow, or move the vehicle to a safer location. The researchers developed 'scenario trees' to map these transitions, demonstrating the heavy cognitive load placed on operators. While technical enhancements could reduce reliance on human intervention, the paper argues that external assistance—such as remote operators or cloud-based decision support—will be indispensable during the ramp-up phase of AV deployment. The findings underscore the urgent need for clearer regulations and more robust technical solutions to ensure safe AV operations.