Optimized Building Component Assessment Planning Using a Value of Information Model

Abstract

Facility condition assessments are a key aspect of a proactive asset management approach, as they provide the information necessary to determine current conditions and work needs, as well as forecasting future conditions to identify long-range work planning requirements. While these assessments can provide a benefit to asset management decision making, they also require resources in terms of time, personnel, and funding. When these resources are finite, as they typically are in facility management, it is important to understand the value that an assessment can provide. Of particular interest to facility managers is determining the optimal time to inspect the components of a building to maximize this value. To achieve this, the paper presents a methodology to calculate the value that building component inspection information provides toward the work planning decision process. The methodology uses a value of information decision tree with a Markov component condition prediction model to calculate the probability of failure. This approach considers the time since the last inspection occurred, the expected condition state of the component, the risk consequence associated with failure, and the cost of risk mitigation related to proactive repair or replacement activities. Results of an applied case study indicate that there is indeed an optimal time since the last inspection that optimizes this value, allowing managers to better allocate scarce inspection resources using this model. In addition, varying inspection types with different levels of detail and accuracy can be analyzed, allowing for a selection of the best type of inspection to perform given the component life-cycle situation, in addition to the inspection frequency.