Loading Robots with a Will to Mill
Many robots that operate milling machines in the metal industry are underutilized. A controller that offers greater movement accuracy could enable them to perform milling work on the side. However, this requires that the controller be able to compensate the forces that deflect the cutter from the set path in real time. Tests conducted with a robot at the Technical University of Munich are showing that Siemens is already close to a solution.
Robots lounging around because they’re waiting for work? They actually exist, including the army of robots that load and unload milling machines in the metal industry. Other than their loading operations, they often have nothing else to do. “If they could just do rough milling work on the side, the entire plant would be much more productive,” says Sven Tauchmann, who heads the Robotics Center of Competence at Siemens’ Digital Factory Division in Chemnitz, Germany. Using the Sinumerik Run MyRobot/Direct Control solution launched last fall, loading robots are already performing grinding tasks. Their movements are precise enough for these rough tasks, “But not precise enough for milling,” says Tauchmann. “It’s not yet possible to compensate the forces that deflect the cutter from the set path on its way through the workpiece.” But Tauchmann, whose Center of Competence specializes in improving robots’ accuracy of movement, is already working on a solution. He’s supported by Siemens’ Corporate Technology, where Dirk Hartmann from the Simulation & Digital Twin technology field is already developing a milling force calculation for him that Sinumerik can use to compensate deflection in real time.
Real-Time Correction of the Milling Path
This is just the kind of challenge Hartmann likes: Real-time calculations and simulations are precisely his professional field. In late 2017, he received eighteen months of extensive funding for his solution idea from the Innovation Fund that Siemens established in 2016 to finance inventive employee ideas. For the implementation on a real robot, he’s being assisted by Florian Schnös, who works for the Institute for Machine Tools and Industrial Management at the Technical University of Munich (TU München). Schnös also attends the doctoral program at the Automation and Digitalization Campus research alliance initiated by Siemens, and he’s chosen the milling robot as the subject of his dissertation. The collaboration between Hartmann and Schnös has already yielded its first results. They’ve recently been able to demonstrate a milling/loading robot whose milling path deviates from the set path by less than 100 micrometers. “For rough milling work, that’s already more than sufficient,” says a satisfied Tauchmann, who travelled to Munich with five other colleagues from the Digital Factory unit to see what the Hartmann and Schnös robot can do.
Real-time Information about the Tool
Tauchmann would like to integrate the milling force calculation into Sinumerik as soon as possible. “That would be a really nice addition to our product line that supports metalworking by robots, from programming the milling paths using NX CAM to controlling them using Sinumerik,” says Tauchmann. “With no additional expenditures, we could turn a loading robot into a milling robot that does rough preparatory work for the milling machine it operates. This would significantly increase its throughput.” But future milling robots aren’t the only reason Tauchmann and Hartmann are interested in the milling force calculations. “Using conventional milling machine controllers, it would be very easy to acquire information on the status of the tool indirectly, by comparing the calculated milling forces with the actual milling forces that the robot registers via its sensors;” explains Hartmann. “A blunt tool that causes inappropriate milling forces would then be detected before damaged workpieces made a tool change necessary.” (fk)
Subscribe to our Newsletter
Stay up to date at all times: everything you need to know about electrification, automation, and digitalization.