Internet of Things-Based Real-Time Thermal Monitoring: Utilizing the Affordable MLX90640 Thermal Camera and the Industrial MAX6675 Thermocouple Module with a Linear Regression Characterization Method

Thermal monitoring is essential for preventive maintenance to ensure that electrical and mechanical devices operate safely within their optimal temperature range. This research developed a temperature monitoring system comprising an MLX90640 thermal camera and four type-K thermocouples, all integrated with the MAX6675 module and connected to the Internet of Things (IoT) system. The characterization of the temperature sensors uses an industrial thermogun to determine their ideal distance and accuracy with the linear regression method. The optimal distance for the thermal camera from the heat source is 15 cm, but the proper distance for the type-K thermocouple sensor at the MAX6675 module is direct contact with the heat source surface. The evaluation of the monitoring system involved assessing the temperature of two distinct items within a temperature spectrum, ranging from 36°C to 48°C and from 36°C to 96°C. The data from the array sensor system indicate a high degree of accuracy in relative sensor measurements, exhibiting few mistakes. The sensor's relative measurement error was relatively small: 1.29% for thermal camera MLX90640, 1.83% for thermocouple 1, 0.9% for thermocouple 2, 1.04% for thermocouple 3, and 1.9% for thermocouple 4. The advance of this IoT-integrated monitoring system is developed using the MING Stack environment (MQTT, InfluxDB, Node-RED, and Grafana), a popular collection of open-source technologies used for Industrial Internet of Things (IIoT) applications. The results indicate that the system can efficiently monitor temperature; the dashboard provides real-time insights and historical data analysis, allowing users to make informed decisions regarding temperature management.