Global path planning for inspection robots in uneven terrain of construction sites

Abstract Manual inspection in construction sites suffers from low efficiency, inconsistent inspection quality, and insufficient digitalization of inspection results, making it difficult to meet the requirements of modern construction safety management. With the rapid development of construction robotics, inspection robots have emerged as an effective solution for on-site inspection tasks, among which global path planning plays a critical role in ensuring safe and reliable robot navigation in complex and uneven terrain. To address the challenges posed by terrain undulations in construction sites, this study constructs a global elevation map of the construction environment and proposes a terrain-aware constrained A* global path planning algorithm. The proposed algorithm incorporates a bidirectional search strategy and a path node filtering mechanism, while explicitly introducing slope constraints and kinematic constraints to enhance path safety and executability. Experimental validation is conducted using both a construction site simulation environment and data collected from a real construction site. The results demonstrate that the proposed method is capable of generating smooth and safe global paths, effectively reducing the risk of robot rollover in complex terrain, and confirm its engineering applicability in real construction site environments.