Development of a Mathematical Model for the Estimation of Required Maintenance for a Homogenous Facilities Portfolio Using Multiple Linear Regression

The maintenance and upkeep of a university facilities portfolio requires the facilities manager to be vigil in overseeing the facilities’ care and to aggressively and innovatively pursue maintenance funds. The problem of insufficient maintenance funds compounded by the budgetary requirement of advanced fund requests in the university environment can cause shortfalls in maintenance dollars. This lack of maintenance funding has led to increased deferred maintenance. Deferring maintenance has negative consequences for the university’s mission. Thus, in pursuing university dollars to further maintenance activities, a facilities manager must be able to substantiate the funds requested.

This research discusses the existing maintenance-prediction models that have contributed in estimating maintenance costs. Also, the causes and impacts of deferring maintenance are investigated in the literature review. The research shows how a facilities manager can take historical facility-attribute data from a maintenance work-order system and develop a prediction equation by using multiple regression analysis for predicting required maintenance. The derived prediction equation’s results were compared with those of three popular models discussed in the research. The prediction equation’s results strongly correlated with all three of the models’