Building Health Monitoring of Hospital Facilities Supported by Digital Twins Models

Emerging Building Information Modelling (BIM) tools and technologies have gradually changed how information about our built environment is created, stored, and exchanged between involved stakeholders. This paper elaborates on viable ways of crossing from a BIM world view with its existing knowledge domain and specific technologies towards a Digital Twin (DT) world view, which promises more significant potential at the intersection of IoT and AI through semantic models. This should address the challenge of changing from its static, closed data with recursive interoperability issues towards a linked data paradigm. The building product can be fully represented in the form of a Digital Twin. The present thesis proposal intends to develop a building health monitoring (BHM) system applied in hospital assets using digital twin concepts based on an optimized method of optimally sensor networks supported by a digital twin. The sensors' initial location will be determined by their ability to detect structural and non-structural influencing parameters, like, temperature, strains or deflections, and dynamic properties, among other paraments base on an initial building model. A digital twin model will receive the sensors network's information, update the building model, and evaluate the building response under the climate actions and extreme events. In the end, the framework will be able to predict the building behavior and the operation and maintenance needs. In this study, by using neural network hierarchy and innovating in damage index as neural network input, the damage detection technique has high efficiency for detecting failures in hospital constructive elements.