Motivation and Goals

The digital twin approach provides a promising support for the challenges faced by the construction industry. However, its application across this sector remains limited. An important obstacle is that traditional digital twins rely on deterministic models, which require unique values of input parameters. This limits accuracy, as such models can not account for the uncertainties inherent in construction projects. These uncertainties are particularly pronounced in geotechnical design and construction. To address this challenge, we propose a Probabilistic Digital Twin framework that extends traditional digital twin methodologies by incorporating and propagating uncertainties from multiple sources.

Recent Results 

After developing the theoretical foundation of the framework (see Figure 1), we implemented a prototype and validated it through a road-embankment case study (see Figure 2). Findings have been summarized in a manuscript recently submitted to the Geodata and AI journal. Additionally, we are investigating with a masters student, a more complex case study involving an excavation pit in an urban setting. Preliminary results of this investigation will be presented at a conference in August 2025, with plans to expand the research to to provide results for a journal publication. In parallel, we are preparing a another journal article comparing 3D subsoil modeling approaches to determine the most suitable method for integration into the probabilistic digital twin framework.