Modeling of specific safety-critical driving scenarios for data synthesis in the context of autonomous driving software

Focusing on the development of autonomous driving technology, this work emphasizes the critical need for precise testing of software components to enhance road safety and reduce traffic accidents. It addresses the challenges of safety-critical driving scenarios, which have not been thoroughly analyzed in terms of forensic evaluation or AI solutions. The use of kinematic models is proposed to simulate vehicle movements under specific constraints, producing synthesized data for training machine learning algorithms. Key scenarios modeled include emergency braking, turning, and overtaking.