Digitalization of underground infrastructure for effective asset management: Technological Approaches, Opportunities and Challenges

As urbanization accelerates, underground infrastructures, including utility networks, have become vital components of cities, providing essential services to urban life. These infrastructures are crucial assets requiring efficient planning, construction, and maintenance. Integrating technologies like Building Information Modeling (BIM), machine learning, and computer vision into underground construction and maintenance enhances visibility and reliability. However, applying these technologies presents unique challenges and opportunities distinct from above-ground contexts. This study therefore adopts a systematic review of 26 articles to discuss the technological approaches, opportunities and challenges in digitalizing underground infrastructure. Key opportunities include integrating GPS, GIS, AR, and computer vision to improve locating accuracy and efficiency. 3D models and advanced visualization systems can prevent excavation damages, while BIM improves construction management and coordination. Accurate defect characterization and infrastructure condition diagnosis through information modeling are also highlighted. Challenges include the limited capabilities of traditional image processing, a shortage of accurate as-built and as-is information, and difficulties in reconstructing buried infrastructure models. Data collection for deep learning is labor-intensive, and automated data interpretation faces uncertainties. Additionally, existing robotic systems are not fully automated, and reliance on historical data affects prediction models. Addressing these challenges is essential for the effective digitalization of underground infrastructure.