An Integrated Electric MotoMonitoring System with Display and IoT Connectivity for Electric Motor Assessment

Abstract

Electric motors are the backbone of industrial automation and manufacturing systems, yet motor failures due to undetected mechanical imbalances, misalignment, or vibration anomalies can result in costly downtime. This thesis proposes the design and implementation of an Integrated Electric Motor Monitoring System that utilizes Condition Monitoring Sensor for real-time condition monitoring. The system features a local visual display interface for immediate on-site diagnostics and an IoT-based data transmission module for remote access and cloud-based trend analysis. The sensor offers precise vibration measurement across multiple axes and is suitable for harsh industrial environments, enabling the system to detect early signs of motor wear or imbalance. The data acquired is processed using a microcontroller, displayed on an HMI (Human-Machine Interface), and transmitted via MQTT protocol to a cloud dashboard for remote monitoring and predictive maintenance strategies. The objective of this research is to improve motor reliability, reduce maintenance costs, and promote preventive diagnostics by integrating sensor-based analytics with IoT infrastructure. This project aligns with the broader vision of Industry 4.0 and smart factory implementations, emphasizing predictive maintenance and system integration. Previous studies, such as "Development of Condition Monitoring System Based on Vibration Analysis for Predictive Maintenance in Industrial Motors" (Siddique, A., Yadava, G.S., Singh, B., IEEE Transactions on Industrial Electronics, 2003), have highlighted the effectiveness of vibration-based monitoring in detecting motor faults. By integrating such sensors with IoT and display technologies, this research extends the conventional model to offer a more scalable, connected, and user-friendly motor health monitoring solution. The implemented system was validated on an operating induction motor. The automatic baseline-teaching procedure, executed from the web dashboard, learned a healthy vibration magnitude of B = 1.09 mm/s and derived the operating limits accordingly (off threshold 0.44 mm/s, pre-warning 1.96 mm/s, and alarm 2.73 mm/s). Using these learned limits, the system correctly classified the OFF, NORMAL, WARNING, and latched ALARM states in real time, remained consistent between the local LVGL HMI and the HTML web dashboard over the WISP-mode MQTT/HTTPS link, and sustained continuous cloud updates at one dashboard read per device update, keeping operation comfortably within the free quota of the cloud service.