Fault Detection in Building Infrastructure Using IoT Sensors and Bayesian Network

Abstract

Optimizing the building performance and component future has never been easier than integrating Internet of Things (IoT) sensors in building infrastructure. This technology has transformed fault detection and prognosis. This research provides a thorough method for defect identification and prognostics using Bayesian network (BN) modeling approaches and IoT sensors. Deploying a network of IoT sensors across the building's infrastructure is the first step in the method. These sensors will continually monitor metrics such as temperature, humidity, energy usage, and equipment health. Anomalies and deviations from typical operating circumstances may be detected by these sensors via their real-time data streams. Second, the data collected by the sensors will be analyzed using BN, which are a probabilistic graphical model to determine the potential of building infrastructure failures. To effectively diagnose and predict failures, BN makes it possible to describe complicated correlations among variables. The proposed method identifies problems, including equipment malfunctions, BVAC system breakdowns, and energy inefficiencies, by merging data from IoT sensors with BN analysis. It can predict when faults will arise, which helps with preventive maintenance and less downtime. It also includes a case study that shows how the method worked in a real-life situation involving building infrastructure, demonstrating how operating efficiency was improved and maintenance costs were reduced.