AIBIZ (CEO Ha Seung-jae), an AI process management company, announced on the 2nd that its 'anomaly detection system' using sensor process data has obtained a domestic patent. The patent is titled 'Method for detecting abnormal operation of process equipment, computing device, and computer program.' When filing in Korea, it also filed patents in the United States, China, and Taiwan. The technology is characterized by technical differentiation from conventional anomaly detection systems by reflecting 'correlations between sensors.' It analyzes with AI the time-series sensor process data acquired during the process. For example, it is common sense that pressure rises when temperature rises in a confined space, but if correlations between sensors are not learned, the accuracy of factory equipment anomaly detection can decline. One example is a case where an anomaly is detected in temperature but the change in pressure is not detected. The patented algorithm finds and learns the correlations between sensors on its own. In particular, engineers can input and train it on what they know well. AIBIZ explained that the anomaly detection results match the engineer's perspective, making defect analysis intuitive. An AIBIZ official said, "It is an anomaly detection technology that can cover the vast amount of sensor process data pouring in real time," adding, "As factory equipment becomes more complex with the introduction of FA (factory automation) systems, AI anomaly detection technology will help maintain factory stability." <Copyright (c) Money Today, 'Real-time news that shows you the money.' Unauthorized reproduction and redistribution prohibited>