NEED for AI:
To enable the development of a machine learning model that can accurately predict tool wear in CNC machines. This model can be used to optimize machining processes, reduce machine damage, and improve the quality of parts produced.
AI REGIO SOLUTION:
Use of federated learning (FL) to address the privacy concerns of sharing sensitive data in the manufacturing industry. The experiment aims to develop a FL model for CNC tool wear by retrieving data from multiple CNCs, training the model on this data, and distributing it to the sources for execution on the edge. This approach allows for collaborative learning without exposing sensitive data, enabling the creation of a shared prediction model while keeping the training data on the device.
Increased digitization level of the company
Reduction in machine failures due to tool wear
Decrease in cost of defective product produced
Broadening of the CNC facility's portfolio by mutual learning, cooperation and sharing of tools and competencies
The AI4CNC experiment is developing a Computerized Numerical Control (CNC) manufacturing data space for one of TEKNOPAR's CNC facilities in Izmir. The project's primary objective is to enable federated learning AI models to estimate the tool wear of CNC machines. The developed solution can be marketed in digital platforms and may also be beneficial with support from AI REGIO.