Advancements in Integrating AI with Geospatial Data for Autonomous Navigation of Land Vehicles

Abstract

Autonomous navigation systems for land vehicles require precise environmental perception, real-time decisionmaking, and robust trajectory planning. This paper presents the development of an autonomous navigation system that integrates geospatial data with artificial intelligence (AI) techniques to enhance localization, obstacle detection, and path planning. The proposed system fuses multiple data sources, including GPS, LiDAR, and inertial measurement units (IMUs), with AI-driven models for sensor fusion, scene interpretation, and adaptive control. Deep learning and reinforcement learning algorithms are employed to optimize navigation in dynamic and unstructured environments. Experimental validation is conducted using a prototype vehicle in diverse terrains, assessing localization accuracy, obstacle avoidance efficiency, and real-time adaptability. The results demonstrate improved navigation robustness and efficiency compared to traditional rule-based methods. This research contributes to advancing autonomous vehicle navigation by leveraging AI and geospatial data fusion to enhance operational safety and performance.