An innovative application of CNN-based real-time temperature monitoring in electronics using thermochromic PLA

The growing demand for more powerful electronic equipment forces the operator to use active control systems such as temperature monitoring. Real-time temperature monitoring is critical for ensuring the longevity and operational reliability of electronic components. This study presents a contactless, cost-effective, and user-independent monitoring system. In an innovative approach, thermochromic Polylactic Acid (PLA) is utilized as a passive temperature sensor. The proposed methodology employs a Convolutional Neural Network (CNN) to determine the temperature gradients of a Gigabyte GA-EQ45M-S2 motherboard and an Intel Core 2 Duo E8500 processor. Experimental results demonstrate that the system achieves 96% accuracy in image-based temperature detection. Measurements were compared with data obtained from a thermal camera and an infrared (IR) thermometer to validate the proposed method, under the same conditions. The findings suggest that this CNN-based system provides a viable, low-cost alternative to expensive thermal imaging equipment and traditional sensors for electronic component real-time temperature monitoring.