A study in the International Journal of Mechatronics and Manufacturing Systems shows how artificial intelligence (AI) can be used to extend the life of robotics components in the manufacturing sector by improving so-called power skiving tools.
Power skiving is an advanced method used in gear production that involves high-speed cutting. The tools used in these processes are highly susceptible to wear and tear and if this wear is not detected early, it can lead to costly replacements and re-sharpening, cutting into the profit margins of the product. Daniel Kiefer, Florian Grimm, Tim Straub, and Günter Bitsch of Reutlingen University, and Clemens Van Dinther of Karlsruhe Institut of Technology, Germany, have now shown that AI can be used to inspect and assess images of the cutting edges of tools acquired by a camera-equipped robotic cell and so provide early warnings to the manufacturing plant of imminent failures.
The AI uses a generative deep learning model that has been trained on images of anomalies that does not rely on a massive dataset of worn tool images. The new approach also side steps the need for images of defect-free tools for its training. The generative model essentially learns from a smaller dataset but can still identify wear patterns with remarkable precision and so predict the remaining useful life of a tool. This will help smaller manufacturers who may not have access to the large, expensive data sets needed by other AI models.
The automation of processes such as tool inspection and wear detection in this way could support workers, allowing them to focus on higher-level tasks. But, it also improves decision-making and minimizes human error.
Kiefer, D., Grimm, F., Straub, T., Bitsch, G. and Van Dinther, C. (2024) ‘Enhancing power skiving tool longevity: the synergy of AI and robotics in manufacturing automation’, Int. J. Mechatronics and Manufacturing Systems, Vol. 17, No. 2, pp.201–224.
No comments:
Post a Comment