We’re excited to add yet another successfully finished project to our list. Over the years, our research department has collaborated closely with renowned research institutes and universities worldwide on over 80 research projects. This time, the team leveraged our extensive experience in AI and machine learning to develop a system that accurately predicts machining times based on CNC part programs and machining parameters.
AI-based machining time prediction
Accurately predicting machining time is crucial for effective production planning and optimization. While Computer-Aided Manufacturing solutions estimate job execution times based on CNC part programs, actual processing times can and often do vary significantly.
To address this challenge and put our AI and machine learning skills to the test, we responded to the KITT4SME Open Call, which invited technical developers to propose solutions that address problems for EU SME manufacturers. With the CNCSmart project, we had the opportunity to collaborate with the KITT4SME partners, gaining access to technical know-how and the chance to create a success story for our solution through pilot experiments. Our role was to develop the complete CNCSmart solution and validate it in real-world environments.
The CNCSmart service predicts the actual execution times of CNC part programs for cutting machines in the metalworking industry. We have developed an AI solution that builds upon existing AI technolology, incorporating novel data sources and incremental machine learning. By analyzing CNC part programs and utilizing historical data, along with various cutting parameters such as machine type (e.g., oxyfuel, plasma, laser cutting), material type and thickness, and operation (e.g., cutting, marking, drilling, piercing, bevel cutting), our system accurately estimates execution times.
CNCSmart serves as a baseline to develop prediction services for any CNC-enabled machinery that provides data through OPC/UA, ADS, or other standards. Our solution is already available in the KITT4SME platform and will also be integrated into our client’s existing proprietary application.
Valuable lessons learned
“The project provided us with valuable insights into the obstacles, challenges, and standards that a robust industrial IT solution must overcome during its development and throughout its lifespan,” explains Nejc Bat, XLAB’s project manager.
We had the pleasure of working with our long-term partner, Messer Cutting Systems (MCS), a leading supplier of thermal cutting solutions for the metalworking industry. MCS provided us with relevant data and access to the necessary IT services for validating the CNCSmart solution.
“Their professionalism and in-depth knowledge of their specific field allowed us to identify and overcome some aspects of the system that we would probably never have discovered ourselves.”
By being part of the KITT4SME project, we gained invaluable experience in how an integration project evolves with novel cloud/container technologies. “Working with both ends of the spectrum provided our team with a wealth of experience that will undoubtedly pay dividends in the future,” concludes Bat.