Abstract: To address the difficulty in monitoring overheating faults of elevator traction machines, this study constructs a monitoring system based on infrared thermal imaging technology. A FLIR T660 thermal imager with a resolution of 640×512 and a temperature measurement range of -20~150 ℃ is adopted to extract 3 characteristic parameters, including the maximum temperature of the stator winding. Combined with Fourier's Law, a heat conduction model is established, and a three-level fault determination rule is designed (it is considered normal when K≤0.3 and T≤0.5 K/min). The experimental results show that when the stator winding is overloaded, the infrared measured temperature is 95.7 ℃, which is 27.5 ℃ higher than the normal temperature, and the K value is 0.82. In addition, when the bearing lubrication fails, the bearing temperature reaches 88.9 ℃, which is 46.4 ℃ higher than the normal temperature. Therefore, the infrared thermal imaging technology can accurately locate faults with a low misjudgment rate.