Markov model based building deterioration prediction and ISO factor analysis for building management

While the Markov chain was successfully used for predicting the average condition of a network of assets under one influencing factor, incorporating the effect of a number of factors into a Markov chain model requires a separate analysis of each factor. In particular, modeling building deterioration is overly complicated due to the complexity in number and hierarchy of building components. The research study reported in this paper aims to combine the Markov chain with ISO factor method-based framework to offer a more reliable deterioration forecasting approach for buildings. This paper identifies major influencing factors for building deterioration. It investigates the ability to use Markov chain for deterioration modeling of a selected critical building component. The building condition inspection data from a local government agency in the State of Victoria, Australia, is used to calibrate a Markov deterioration model considering a number of influencing factors. To help demonstrate the concept, two factors influencing deterioration are considered in the analysis: outdoor environment and in-use condition. Validations of the models are undertaken using a subsequent inspection data set using Pearson's chi square test. A generic process for deriving Markov chain-based deterioration curves from discrete condition data incorporating factors influencing building deterioration is presented in this paper. Validated models and the method can be applied to reliably predict the deterioration of buildings with complex hierarchy, which gives a step change in building management.