Failure Diagnosis and Maintenance of Industrial Robot Based on Deep Learning and Augmented Reality

To produce products within a customer-tailored manufacturing system, conventional industrial robotic equipment handles various processes such as assembly and inspection. However, depending on the customer's needs, such as orders and requirements, the types and specifications of products may change. Moreover, some production processes may be added or omitted, and the types and materials of parts used in the same process may change. Accordingly, workers should recognize the changed situation of the manufacturing system and take appropriate measures. However, workers cannot recognize and cope with all the changes in the production process or situation of the manufacturing system, and they may handle unexpected failures of the industrial robot-ic equipment poorly, impairing work efficiency. Therefore, workers must be able to obtain timely and utilize effectively appropriate management information on the problems of the industrial robot equipment required to perform specific tasks and appropriate maintenance in order to improve the productivity of the manufacturing system and realize value creation such as increased profits and reduced costs. This requires fault detection technology of industrial robotic equipment to find problems in robotic equipment in a timely manner, and necessary maintenance and management information should be provided according to the worker's environment, such as tasks, roles, and abnormal conditions.