Digital twin for cultural heritage: A computational approach to predictive conservation

The shift towards predictive and intelligent management models in cultural asset conservation has gained significance due to the intricacies of degradation processes and the necessity for sustainable preservation solutions. This project aims to develop a methodological framework for creating a Digital Twin (DT) for architectural heritage, which can integrate geometric, historical, environmental, and predictive data into a cohesive, dynamic system. The suggested method integrates high-accuracy surveying techniques with semantic modeling using Heritage Building Information Modelling (HBIM), augmented by real-time data collection using Internet of Things (IoT) devices. Environmental and structural parameters are perpetually monitored and integrated with the digital model using visual programming procedures, facilitating real-time changes and interactions. Machine Learning (ML) techniques are employed to analyze time-series data for the identification of deterioration trends and the simulation of predictive maintenance scenarios. The technique was validated by its application to the Ponte Leproso, a Roman bridge in Benevento, Italy, noted for its intricate stratifications and susceptibility to environmental stresses. The development of a DT of the structure facilitated the dynamic integration of sensor data with historical and architectural knowledge, hence enabling the formulation of data-driven conservation plans. This integrated workflow illustrates how the collaboration of HBIM, IoT, and AI technologies may facilitate the transition of cultural heritage management from reactive intervention to proactive, intelligent, and sustainable preservation methods.