A Digital Twin For Deep Vibro Ground Improvement

This paper presents the implementation of a digital twin for the vibro-replacement ground improvement process, as part of a digital construction site. The goal is to implement an evidence-based quality assurance for each produced element (column or point). The system uses a combination of planning data and real-time machine data collected automatically from construction equipment. In addition to the quality assurance, feedback is provided in a number of areas, e.g., condition monitoring of the equipment. The digital twin builds upon the real-time sensor and machine data, fused with metadata, to implement a digital model for the process. A hierarchical data evaluation, using rule-based Key Performance Indicators (KPIs), permits viewing of the spatial variation of data over the construction site. This permits the modelling of systematic variations in subsurface properties across the site, enabling the detection of anomalous individual elements, the diagnosis of their cause and the prediction of properties as construction proceeds over the site. All time-series data and KPIs are represented by standard classes of objects, leading to a generic coding of the data analysis; this minimizes the effort required to add additional sites and/or new KPIs. Examples are presented for a number of sites, where the system was used to identify anomalous columns.