Improving the Performance Prediction of Process Simulation Models for TBM Tunneling Using Real-Time Project Data

The planning of production, logistics, and maintenance in mechanized tunneling relies on project-specific conditions and assumptions about numerous production parameters. Analyzing multiple scenarios for TBM production is crucial for ensuring robust project flow. Process simulation models are valuable for this analysis. However, deviations from the plan often occur due to uncertainties in initial assumptions and unforeseen events during construction, especially given the variable ground conditions in TBM tunneling. This paper presents a concept for adapting offline simulation models used in the planning phase for real-time optimization by integrating actual project data. Continuously updating these models with new data from the construction site improves performance predictions. The concept involves integrating current construction progress into the simulation model and updating the originally assumed input data with project-specific actual data. An example study demonstrates the continuous adjustment of a simulation model and its impact on project duration prognosis. The enhancement of the prognosis by integrating real-time data is investigated, and an initial assessment of the earliest possible time for an update is made. The benefits of using real-time data in process simulation during construction are discussed, highlighting the improved accuracy in performance prediction and the potential for more efficient project management.