ABSTRACT
Mechanised tunnelling in soft ground whether in densely populated cities, under/vicinity of buildings or green field areas and involving large volumes of excavation, is likely to pose a major risk of surface anomaly from over excavation during tunnelling process. The severity of the damage can be significantly large when it comes to large diameter tunnels. Investigations of incidents often blame ground conditions, however there are a number of other factors that are contributing or even the root cause, such as insufficient ground investigation, lack of pre-tunnelling assessment, implementing inappropriate tunnelling methods, inadequate or timely operational control or ineffective monitoring systems etc. In Singapore large TBM tunnelling is at its early stages. The focus of this paper is on making quick changes to TBM operations with rapid reaction time and avoid incidents. It is based on a holistic assessment during tunnelling using live TBM data using machine data regression analysis including analysis of operational efficiency. Further, it suggests how artificial intelligence and machine learning algorithms can help as additional tools to enhance quick decisions.
