Exploring the Boundaries: Thorough Software Testing for Safety-Critical Driving Scenarios Based on Kinematics in the Context of Autonomous Driving

Résumé : This scientific publication focuses on the efficient application of boundary value analysis in the testing of corner cases for kinematic-based safety-critical driving scenarios within the domain of autonomous driving. Corner cases, which represent infrequent and crucial situations, present notable obstacles to the reliability and safety of autonomous driving systems. This paper emphasizes the significance of employing boundary value analysis, a systematic technique for identifying critical boundaries and values, to achieve comprehensive testing coverage. By identifying and testing extreme and boundary conditions, such as minimum distances, this publication aims to improve the performance and robustness of autonomous driving systems in safety-critical scenarios. The insights and methodologies presented in this paper can serve as a guide for researchers, developers, and regulators in effectively addressing the challenges posed by corner cases and ensuring the reliability and safety of autonomous driving systems under real-world driving conditions.