Analysis of Using Machine Learning to Enhance the Efficiency of Facilities Management in the UAE

Abstract

The UAE's built environment industry faces challenges due to energy consumption trends and inefficiencies in Facilities Management (FM), exacerbated by traditional maintenance methods relying on outdated paperwork and limited technological integration. This study addresses these issues by Implementing Machine Learning (ML) algorithms using data from Building Management Systems (BMS) and FM maintenance reports, focussing on predictive maintenance for Fresh Air Handling Units. Seasonal thresholds are set for sensor values, and the algorithm detects hazardous trends when values exceed safe limits, correlating them with related sensors to generate combined variables and error codes. Each error code represents specific trends and prescribed actions to address potential malfunctions. The approach is validated using linear regression and mathematical logic, while binary logistic regression predicts issues with blower motors and filters by analyzing BMS data. A comparison between traditional and ML-based FM strategies demonstrates that ML-driven scheduling and maintenance reduce costs by 38% to 50%, significantly improving efficiency. This study highlights the transformative potential of ML in modernizing FM practices, enabling proactive, cost-effective, and efficient maintenance operations in the UAE's built environment sector.