Challenge
Our client operates a production facility for metal components, where anomalies in the machine ejection of produced parts cause disruptions in the production process. The goal was to detect these anomalies in real-time and stop production if necessary to prevent damage to the production machinery and increase production speed.
Approach
We developed a camera-based AI system that detects anomalies in real-time. We trained the model in the cloud and used machine data as well as image data for the prediction. Our approach included:
• Using AI-powered image recognition and machine data for error detection in the production process
• Immediate response and production stoppage upon anomaly detection
• Local implementation of the AI system in the production facility to minimize latency
Results
The implementation of our camera-based AI system led to significant outcomes:
• Real-time error detection in the production process through AI-powered image recognition
• Increased production speed
• Prevention of machine damage
• Local operation of the AI system within the production facility
In future, the camera and software will be used for another machine in order to map other aspects of automated image processing with a new model.