Abstract: Disclosed are an equipment malfunction diagnosis apparatus using a sound spectrogram image, and a method therefor. In other words, the present invention collects sound data through a sensor unit provided at one side of equipment, converts the collected sound data into a spectrogram image, and determines malfunction of the equipment and classifies a cause of the malfunction by performing machine learning using the spectrogram image as an input value, thereby quickly and accurately identifying the cause of the malfunction when it is identified that there is the malfunction in the equipment, and providing a counterplan in response thereto.