<![CDATA[Early identification of abnormal temperature conditions is important for preventing overheating from developing into more serious thermal hazards. Thermal array sensors provide a non-contact approach for observing temperature distribution; however, their measurement behavior may vary with sensor-to-object distance and environmental conditions. This study experimentally evaluates the measurement performance of the AMG8833 thermal sensor for hotspot detection at different sensor-to-object distances and under different environmental lighting conditions. Measurements were conducted at distances of 5, 10, 15, 20, 30, 40, and 50 cm under bright, dim, and dark conditions. Each experimental combination was repeated five times, and one representative measurement was selected based on the relative stability of the temperature reading and hotspot pattern during the measurement process. The AMG8833 measurements were compared with a reference thermal gun to determine the measurement difference. The results showed that the mean measurement difference across the three conditions increased from 0.60°C at 5 cm to 2.73°C at 50 cm, with an R2 of 0.956 for the linear fit of mean difference versus distance. The mean measurement differences across all tested distances were 1.59°C, 1.36°C, and 1.51°C under bright, dim, and dark conditions, respectively. Using a practical criterion of a mean measurement difference of ≤1°C over the evaluated interval, the 5–20 cm interval was identified as the practical operating range for the tested hotspot measurement configuration. These results indicate that sensor-to-object distance produced a more pronounced variation in measurement difference than the tested environmental lighting conditions.]]>
