Abstract

In the protection of heritage buildings, computational models are an innovative technology for the implementation of effective preservation strategies. These models can manage both data obtained from the buildings and data from the knowledge base of professional experts. The preventive conservation of heritage buildings requires cooperative efforts, where the professional experts' opinions are paramount, to analyse multidisciplinary knowledge. In this respect, the incorporation of new protocols, which can help decision makers prioritize interventions and avoid restoration actions that are unnecessarily invasive or might cause irreparable loss of the properties, is urgently needed. The aim of this study is to establish a new approach concerning a digital management system that is based on artificial intelligence (fuzzy logic). Application of the model reveals that the sample (42 buildings) was classified into two vulnerability conditions: (i) medium-level affections and (ii) high-level vulnerability affections. The results obtained in this study will be useful to stakeholders who are responsible for the maintenance of heritage buildings, since this methodology allows reducing their probability of failure. This new digital management tool will help in the establishment of possible mitigation strategies that are focused on preventive future maintenance programmes of a set of heritage buildings in South Chile.