There have been a lot of studies in the past for the method of predicting the failure of a machine, and recently, a lot of researches and applications have been generated to diagnose the physical condition of the machine and the parts and to calculate...
There have been a lot of studies in the past for the method of predicting the failure of a machine, and recently, a lot of researches and applications have been generated to diagnose the physical condition of the machine and the parts and to calculate the remaining life through various methods. Survival models are also used to predict plant failures based on past anomaly cycles.
In this paper, we developed a rail-driven mobile system that diagnoses machine abnormalities by receiving machine information from images and thermal images of targets with limited operator access and inspection or requiring long-term monitoring in a wide area. This anomaly diagnosis model uses a 2D temperature array (temperature map) and a motion vector array (motion map) as inputs data of the Wasserstein GAN (wGAN) learning model trained with normal data. Because the difference between the input and output is used as a global loss function (global), the anomaly score is calculated and the condition of the machine is diagnosed according change of anomaly score. This rail-driven mobile system is expected to make a great contribution to equipment maintenance and optimal operation because it provides quantitative evidence to monitor equipment in the long term to diagnose failures early and give maintenance instructions.