Safety of the Intended Functionality

Safety of the Intended Functionality (SOTIF) focuses on ensuring that automated systems, particularly in autonomous driving, behave as intended and do not pose unreasonable risks, even when encountering unexpected situations or system limitations. Current research emphasizes identifying and mitigating functional insufficiencies (FIs) in algorithms like Model Predictive Control (MPC) and perception systems, often using simulation and real-world data analysis to characterize these insufficiencies and their triggering conditions. This work is crucial for validating the safety of autonomous systems and informing the development of robust, reliable, and safe automated technologies, ultimately impacting the commercial deployment and public acceptance of these systems.