Professor Mike Mooney
Grewcock Chair Professor of Underground Construction & Tunneling
Colorado School of Mines
Application of AI to Improve TBM Operations
Artificial intelligence (AI) techniques are making a significant impact in many industries, including underground construction. The application of AI data science/machine learning to improve TBM tunnelling operations requires a significant amount of information to train robust performance prediction models, e.g., for ground detection, tunneling-induced settlement, productivity, tool wear, anomalous behavior, etc. This presentation summarizes an AI framework applied to the Washington DC North-East Boundary Tunnel project and to other ongoing projects in North America. The presentation introduces aspects of TBM tunnelling that can be positively impacted by machine learning, including ground characterization, tool wear estimation, advance rate prediction and tunnelling-induced deformation. Results from AI application in the US show promising results, much to learn and significant promise for AI adoption in underground construction.