Machine learning techniques based data-driven fault detection of multiple transducers

Monitoring the state of the facilities and detecting any technical defects in the smart factory is critical. As the industry advances, technical systems get more advanced. The more complicated the system, the more prone it is to failure. To protect a system once it has become flawed, flaws must be swiftly discovered, isolated, and repaired, this necessitates the use of effective Fault Detection and Isolation (FDI) procedures. Corrective steps must be conducted immediately if the sensors and actuators fail. In this paper, an intelligent transformer synthesis fault detection and monitoring of the baby incubator system depending on Machine Learning (ML) techniques have been built. Four machine learning algorithms have been used, Artificial Neural Network (ANN), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Decision Tree (DT) are some of the algorithms used. The performance was evaluated using a fake dataset containing typical characteristics of condition monitoring data collected from the hospital’s neonatal incubator. The experiential results indicate that the use of smart algorithms helps in quickly discovering faults in the system and with high accuracy of up to 100% in the ANN and KNN algorithms.