Improving Autonomous Vehicle Reasoning with Non-Monotonic Logic: Advancing Safety and Performance in Complex Environments

The software design of autonomous vehicles (AVs) incorporates artificial intelligence (AI) characteristics to enhance their safety and overall driving performance. Central to vehicle's operation is the ability to reason effectively in complex and uncertain environments. However, traditional logical systems, such as monotonic logic, often struggle to handle the inherent uncertainties and exceptions encountered in real-world scenarios. This paper proposes the utilization of non-monotonic logic in order to enhance the reasoning capabilities of autonomous vehicles. By incorporating non-monotonic reasoning, vehicles can navigate intricate traffic scenarios, make plausible inferences, and adapt their decisions when faced with conflicting information. This research aims to provide a comprehensive review of non-monotonic logic's application in autonomous vehicles, highlighting its advantages over traditional logical systems and its potential impact on safety and performance. Additionally, through this research, we seek to contribute to the advancement of autonomous driving technology by enhancing the reasoning capabilities of vehicles in various scenarios such as car following related to critical safety events. To overcome the limitations of previous works in car-following models, particularly in their assumptions of homogeneous drivers, the personalized cognitive agent is proposed. The personalized cognitive agent is incorporating heterogeneous driving behaviors, based on individual user preferences, characteristics, and needs. The goal of a personalized cognitive agent is to provide tailored and customized experiences to cognitive vehicles, taking into account the unique requirements and individual preferences of occupants inside autonomous vehicles.