Autonomous Navigation for Unmanned Ground Vehicles in Unstructured Terrain

Abstract

The problem of autonomous navigation for Unmanned Ground Vehicles (UGVs) in unstructured environments is both challenging and crucial for their deployment in real-world applications. Perception is important, as it provides the necessary information for terrain traversability and environmental awareness. In this article, the developed manned-unmanned vehicle designed to carry out autonomous missions in unstructured terrain is presented, along with the system requirements essential for such operations. Challenges related to environmental perception and navigation in unstructured environments are discussed. Achievements in developing AI models capable of interpreting sensor data are highlighted, demonstrating significant progress in the field of autonomous navigation. However, several gaps remain, particularly in the areas of sensor fusion, real-time decision-making, and adaptability to highly dynamic conditions. Development in the field of autonomous systems allows for a wider expansion of the potential applications ofUGVs in various fields, including disaster response, environmental monitoring, and exploration of hazardous environments.