A Framework for Developing Machine Learning Models for Facility Life Cycle Cost Analysis through BIM and IoT

Summary

A large amount of resources are spent on constructing new facilities and maintaining the existing ones. The total cost of facility ownership can be minimized by focusing on reducing the facilities life-cycle costs (LCC) rather than the initial design and construction costs. This thesis presents a research project that developed a machine learning-enabled facility life-cycle cost analysis (LCCA) framework using data provided by Building Information Models (BIM) and the Internet of Things (IoT).

First, a literature review and a questionnaire survey were conducted to determine the independent variables affecting the facility LCC. The potential data sources were summarized, and a data integration process introduced. Then, the framework for developing machine learning models for facility LCCA was proposed. A domain ontology for machine learning-enabled LCCA (LCCA-Onto) was developed to encapsulate knowledge about LCC components and their roles in relation to sibling ontologies that conceptualize the LCCA process. After that, a series of experiments were conducted on a university campus to demonstrate the application of the proposed machine learning-enabled LCCA framework. Finally, the author’s vision of the future smart built environment was discussed.

This research contributes to the body of knowledge by investigating the feasibility of forecasting facilities’ LCCs by implementing machine learning on historical data. By exploring the new possibility for better prediction of a facility’ LCC through leveraging historical data housed in heterogeneous building systems across a continuous network of buildings, this research has a greater impact than simply studying the LCC of an individual project in the design phase. The impact involves data-based LCC inputs in future facilities thus enabling cost benchmarking and informing project developments based on owned historical data. Using existing available data to benchmark facility costs can assist decision making, and new data can be incorporated as they become available. It is an iterative knowledge accumulation of facility costs that could not only identify performance trends and operation and maintenance expense “hot spots”, but also identify the best practices of facility design, construction, and operation from a cost efficiency perspective.

---