Continuous Improvement of Criteria for Condition-based Maintenance by Means of Effects Evaluation of Treatments

With the advancement of sensor and network technologies, condition-based maintenance (CBM) is applied to various industrial equipment. In particular, mechatronics systems are suitable for CBM because of the sensors equipped in the systems. However, criteria for executing preventive maintenance should be properly determined by considering the tradeoff relations between false detections and oversights of symptoms, because the data from the sensors does not directly represent the component deterioration. In our previous paper, we proposed a method for determining the optimal criteria based on operation and maintenance data before applying CBM. This paper proposes a method to improve the criteria based on the data obtained while CBM is implemented. The method is applied to automatic ticket gates for verifying its effectiveness.