OPTIMIZING PROJECT MANAGEMENT EFFICIENCY THROUGH DIGITAL TWIN-ENABLED PREDICTIVE MAINTENANCE IN SMART BUILDING FACILITY MANAGEMENT

Abstract

This research emphasizes designing and testing an intelligent Digital Twin framework that will increase project management efficiency and operational productivity within smart building facility management. The main objective of the system will be to monitor and predict, maintain, optimize energy consumption, and make decisions on energy use. To achieve these objectives, the research method incorporated an end-to-end system design and development process involving architectural modeling, multi-sensor implementation, and data acquisition, with an emphasis on developing an intelligent framework. To build and develop the Digital Twin framework, Message Queuing Telemetry Transport, Hypertext Transfer Protocol, and Modbus were implemented. PostgreSQL, InfluxDB, and SQLite databases were implemented based on data structure and functionality. To develop an end-to-end workflow automation platform, Node-RED programming tools were incorporated. A dynamic interface will be used for visualization and display, which will be directly linked with Power BI. A Long Short Term Memory neural network will be implemented for forecasting and anomaly detection based on energy and environmental data. The findings revealed major enhancements in operational response, situational understanding, and prediction capabilities. The system effectively identified anomalous readings like carbon dioxide levels beyond 600 parts per million, temperature beyond 18-22°C, and humidity beyond 50-80 percent. The Long Short-Term Memory network had a Root Mean Square error value of 0.04, thus signifying efficient prediction. The Digital Twin additionally facilitated better maintenance planning and optimal energy utilization. Overall, the research work represents a scalable and holistic approach that marks progress in intelligent facility management. It integrates sensing, prediction, and visualization capabilities into a single operating environment.