A comprehensive review of control and guidance strategies for unmanned ground vehicles in lane tracking and leader-follower applications

Introduction/purpose: Unmanned Ground Vehicles (UGVs) offer significant advantages for various operations; yet their autonomous control and guidance present substantial difficulties, especially for diverse locomotion types (e.g., tracked, wheeled) in challenging terrains due to complex dynamics, nonholonomic constraints, and environmental interactions. This paper provides a comprehensive review of control and guidance strategies for UGVs, with a specific focus on leader-follower and lane tracking with obstacle avoidance applications. It aims to synthesize the state of the art, identify key challenges generic to UGV autonomy in these tasks, and discuss promising guidance and control methodologies. Methods: An extensive literature review was conducted, analyzing existing research on UGV, autonomy levels, system architectures, control methodologies (including classical, adaptive, robust, and intelligent approaches), guidance approaches, and specific application domains. Methodologies for guidance and control relevant to UGVs in leader-follower and lane tracking tasks were critically examined. Results: The review identifies dominant trends, including the increasing use of deep learning for guidance perception and growing interest in robust control techniques capable of handling UGV operational challenges. Significant challenges persist in perception for unstructured environments, accurate dynamic modeling for diverse UGV platforms, seamless integration of perception with robust control and guidance systems, and extensive real-world validation. Conclusions: Achieving robust autonomy for UGVs in complex real-world scenarios requires integrated solutions addressing guidance and control. Advanced robust control methods emerge as strong candidates for UGV control, but their full potential necessitates further research into their integration with advanced guidance systems.