Discussion about the use of Bayesian networks models for making predictive maintenance decisions

Abstract

The performance of a building decreases with time and this process is accelerated if proper maintenance is not carried out. This paper presents a discussion about the importance of predictive maintenance actions to enhance the condition of existing buildings. An approach based on Bayesian networks (BN) is proposed to predict the condition of a building. The proposed approach consists of a conceptual and generic model, including the factors with more influence in the condition of a building, which were identified by a literature review. The relationships between these factors and a discussion about the application of this model in maintenance decision-making are provided.