Advances in path planning for spherical tank inspection

Abstract

Spherical tanks are widely used pressure special equipment in the petrochemical, energy, and metallurgical industries, and their regular non-destructive testing is a critical link to ensuring safe operation. Path planning technology directly determines the coverage completeness, operational efficiency, motion safety, and trajectory accuracy of inspection robots, serving as the core technology for achieving automated and intelligent spherical tank inspection. This paper presents a systematic review of the research status, typical method classification and applications, technical challenges, and future prospects of path planning in spherical tank inspection scenarios.First, the mobile mechanisms of inspection robots are reviewed, along with key issues such as spherical tank surface constraints, full-coverage requirements, obstacle avoidance constraints, and inspection efficiency, as well as the development of robot engineering applications. Second, existing planning methods are categorized into adaptive path planning methods, intelligent optimization algorithms, multi-robot collaborative planning methods, and deep reinforcement learning approaches, with a comparison of their applicability conditions and research progress. Finally, current research challenges are identified, including high inspection environment costs, complex modeling, insufficient safety and reliability, limited computational resource constraints, and a low degree of integration in end-to-end planning and control. Future research directions are also discussed, such as adaptive planning, multi-constraint optimization, multi-robot collaboration, and the integration of digital twins with deep reinforcement learning. This paper serves as a reference for research related to path planning for intelligent spherical tank inspection robots.