Table of contents:
Video: Artificial Intelligence Detects Wear In Ball Screws
2023 Author: Hannah Pearcy | [email protected]. Last modified: 2023-05-24 11:12
With ball screws, such as those used in lathes for precision guidance in the manufacture of cylindrical components, wear has so far been determined manually. To replace defective components, the machine must be at a standstill. Researchers at KIT have now developed a system for the fully automatic monitoring of ball screws in machine tools that is intended to reduce machine downtime.
“Our approach is based on the integration of an intelligent camera system directly into the ball screw drive. This allows a user to continuously monitor the status of the spindle. If there is a need for action, he will be informed automatically,”explains Professor Jürgen Fleischer from KIT's Institute for Production Technology (WBK).
Artificial intelligence interprets wear precisely
With the new system, a camera with lighting is attached to the nut of the ball screw. As the nut moves on the spindle, it takes individual pictures of each spindle section. This means that the entire surface of the spindle is analyzed.
The captured image data are then evaluated by an artificial intelligence. Users thus receive a direct assessment of the state of the spindle surface.
"We trained our algorithm with thousands of recordings so that it can now confidently distinguish between spindles with and those without a defect," says Tobias Schlagenhauf from WBK, who worked on the development of the system. A further evaluation of the image data quantifies and interprets the wear precisely. "This enables us to differentiate whether discoloration is simply dirt or harmful pitting," explains Schlagenhauf.
With the online machine learning course - how does the machine learn? delve deeper into the methodology. Professor van der Smagt - Head of Artificial Intelligence at Volkswagen - explains to participants in two hours the most important concepts of machine learning in a concentrated and compact manner.
When training the AI, all conceivable forms of visually visible degeneration were taken into account and the functionality of the algorithm was validated with new image data that the model had never seen before. The algorithm is suitable for all applications in which image-based defects on the surface of a spindle are to be identified and can also be transferred to other applications.