Enhancing power skiving tool longevity: the synergy of AI and robotics in manufacturing automation

In gear manufacturing, the longevity and cost-effectiveness of power
skiving tools are essential. This study presents an innovative approach that
combines artificial intelligence and robotics in manufacturing automation to
prevent tool breakage to improve the remaining useful life (RUL). Using a
robotic cell, the system captures six images per tooth from different angles. An
unsupervised generative deep learning model approach is used because it is
more suitable for industrial application as it can be trained with a small number
of defect-free images. It is used in a first step as a classifier and, in a second
step, to segment tool wear. This approach promises economic benefits by
reducing manual inspection and, through automated tool inspection, detecting wear earlier to prevent tool breakage.