Integrating BIM and Bayesian networks to support the management of large building stocks

This research focuses on the development of a decision support system to assist the management of large building stocks with limited budget. The ultimate goal of the study is to rank the priority of buildings refurbishment actions. Owners of any large building stock are usually bound to manage a huge variety of buildings, which very often includes historic buildings and facilities built at different times. Refurbishment of existing building stocks, both to comply with the new regulations and to do organized maintenance, is a complex process that requires an extensive knowledge of constructions that usually involves long times and high costs. A crucial point in the management of large building assets is giving priority to refurbishment and upgrading operations. Due to the nature of existing building and the lack of information, technicians must invest a huge asset and a long time in reaching to achieve a deep knowledge of buildings conditions. Support tools, based on Bayesian networks, help specialists to manage with little knowledge information through the use of probability to handle math uncertainty. Lastly it is possible to weight different evaluation criteria and obtain the complete ranking of the buildings through a multi-criteria analysis. Deducing from obtained scores, we can understand that the higher the score, the more important the need for intervention. This evaluation has the ability in helping owners in prioritizing necessary refurbishment actions.