Challenge
The line between process reliability and time-optimized production is thin. Only a lot of appropriate experience and routine in practice can help. How about making this wealth of experience from long-time users available to every operator?
Thanks to a well-thought-out combination of data analytics and machine learning, new possibilities arise.
Application
Optimized CAD/CAM system
When programming workpieces, manual adjustments are sent anonymously to Analytics. We use the verified data to train a model. Now, the experiences of users worldwide are available to everyone else as valuable input. Analyzing parts with the model is particularly useful in batch processing of large numbers of parts, allowing even experienced users to identify where adjustments are still necessary for process-reliable and efficient production.
TwinAssistant
A retrofittable component is capable of detecting and comparing various objects and thus machine configurations using real-time camera images. The detected components and their positions can then be forwarded to a digital twin. The combination of machine learning algorithms, DataMatrix codes, and mcs expertise are the core elements and success factors of this innovation.
Advantages
Efficiency
- In NC programming, you can only focus on potentially problematic parts.
- The current machine configuration does not need to be updated manually.
Process reliability
- The target configuration can be validated before the process starts.
- Forgotten objects in the engine room are detected.
- Machine downtimes are minimized.
Quality
- Current machine configuration can be compared with proven configurations.