Automated path‐planning strategy for robotic inspection of underground utilities based on building information model

作者：Zihan Yang, Jiangpeng Shu, Jishuang Jiang, Wentao Han, Yichang Wang, Liang Zhao, Yong Bai · 发表于：Computer-Aided Civil and Infrastructure Engineering · 年份：2025 · DOI：10.1111/mice.70107 · 被引用次数：23 · 研究领域：Underground infrastructure and sustainability、3D Surveying and Cultural Heritage、Infrastructure Maintenance and Monitoring

This paper proposes a fully automated end-to-end inspection-path-planning strategy for underground utilities, such as pipelines, based on building information modeling (BIM). An automatic extraction method is developed to process utility information from BIM models, using a registration step that pairs each pipeline with its corresponding utility branch. This is followed by geometric modification via offset algorithms that account for obstacle dimensions to generate safe navigation paths. A novel inspection algorithm, the utility-Chinese postman problem (U-CPP), is introduced to generate a topological map and ensure full-coverage inspection. A Dynamo prototype integrates all these algorithms, minimizing manual intervention and achieving full-process automation. The method is validated with three real-world utility BIM models featuring diverse cross-sectional configurations. The U-CPP algorithm achieves 100% coverage with minimal repetition rates and computes optimized inspection paths in 24, 23, and 23 ms. Results demonstrate that the proposed strategy efficiently automates both information extraction and full-coverage path planning. The U-CPP algorithm proves to be robust, computationally efficient, and effective in handling diverse utility configurations.