Drives and Controls January 2023

34 n GEARS AND GEARBOXES January 2023 www.drivesncontrols.com Digital twins will allow real-time monitoring of gearboxes G earboxes used in robots usually operate with fixed maintenance intervals. After a defined period of time or operating life, they are replaced – regardless of their actual state of wear. Depending on the application and the period and scope of use, however, the gearboxes can be subjected to varying levels of stress. The loads on individual axes varies as well. In these conditions, predictive maintenance has many potential attractions. The Japanese cycloidal gear specialist Nabtesco, which claims to be the world market-leader in robotic gears, is developing an add-on for robots that will provide gear-specific data and characteristics. The aim is to offer digital twins of the gears including their relevant properties. The company predicts that real-time monitoring of robots will, in the future, rely on real digital gear characteristics on the basis of the serial number, allowing a snapshot of the robot load to be generated, and thus an analysis of the remaining life of the gears. The system will make us of data that is already available – speed and torque characteristics, for example – in combination with empirical data from more than 35 years of gear development. It will be able to make detailed inferences about the state of gearboxes and their remaining lives. Constant monitoring of cycloidal gears will facilitate maintenance on demand. It will also allow timely identification of impending failures, while facilitating planning of service calls, reducing maintenance and eliminating unplanned machine downtime. The results, the reserchers predict, will be more efficient production processes and increased productivity. The software-based approach is said to offer numerous advantages over using physical sensors, including lower costs, reduced system complexity and improved data quality. In addition, not all data – for example, rigidity values – can be acquired using sensors. Because gearboxes undergo extensive testing before being sold, Nabtesco says it knows exactly how a gearbox will behave as a result of wear, for example, or how it will be affected by changes in temperature and lubricants. But these parameters will need to be available to robot controllers. Nabtesco is planning to offer digital services that will allow it to monitor digital twins of gears via robot controllers, thus opening up new maintenance concepts in the context of Industry 4.0. Currently the digital twin technology is still being developed. Nabtesco is planning two phases: first, it will develop an add-on with generalised data about particular gear models. And in a later upgrade, it is promising even higher accuracies by using data from gearboxes that have actually been installed. Nabtesco has entered also into a cooperation with the Austrian robot controller expert Keba to develop economical predictive maintenance concepts based on virtual sensors. Digital twins of robot gearboxes will make it possible to monitor the actual operating behaviour and the specific load of each robot axis, and to calculate in real time how much wear there is and whether servicing is required. n A Japanese specialist in gearboxes for robots is working on a project that will allow real-time monitoring of these gearboxes using digital models of the gears, thus delivering predictive maintenance capabilities that will cut costs and boost robot availability.