A Study on the Types of Sensors for Machine Anomaly Analysis and Their Effectiveness

Machine failures can lead to major losses, such as production line shutdowns and quality degradation. Research into failure prediction and anomaly detection is being conducted widely, driven by the need to know the signs of equipment and machinery failure before they occur. This method uses sensors to collect machine status data and use that data to detect signs of failure. It is hoped that the use of AI will enable objective, highly accurate failure prediction and anomaly detection. Sensors are crucial here. Depending on their installation location, there may be issues with the reliability of the data obtained, and signal lines such as power and LAN lines may get in the way of the machinery, making them a key component when considering actual system operation. This paper reports on the results of a study focusing on sensors.