The Karlsruhe Institute of Technology (KIT) is hoping to simplify maintenance of ball screws with its development. These components provide precise guidance of lathes, for example, in the production of cylindrical components. Until now, wear has had to be checked manually. During this time, it was essential to shut down the machine. With the system from KIT , continuous monitoring of the condition of the spindle is now possible.

To achieve this, the researchers fit a camera with a light to the nut of the ball screw. As the nut moves on the spindle, the camera takes individual photos of each spindle section, resulting in an image of the entire surface. The image data is then transmitted to an artificial intelligence (AI) system, which has previously been trained using methods of machine learning and thousands of photos of defective and intact spindles. Using this data, the system can reliably evaluate the condition of the spindle and trigger an appropriate warning if it is worn.