Use cases:
- Anomaly detection
- Defect detection / classification
- Continuous process monitoring
Inspection
Automated inspection is key to ensuring efficient quality control. Whether acquired through cameras, hyperspectral sensors or other metrology hardware, large amounts of data must be analyzed in real-time to avoid production delays. Deep Learning provides better accuracy by learning what features matter directly from your data, and can be cheaply deployed with high throughput on conventional hardware.
Use cases:
- Anomaly detection
- Defect detection / classification
- Continuous process monitoring
Use cases:
- Provide deep insights
- Digital twinning
- Predictive maintenance
Modeling
Predicting the dynamics of a process is the essential first step toward its optimization. Unfortunately, most modern industrial processes are simply too complex to be fully captured by conventional simulation approaches, leaving potentially critical details un-modeled. Our approach is to let the process speak for itself: data sampled from the actual process dynamics is used to learn a predictive model that is bespoke to that application.
Use cases:
- Provide deep insights
- Digital twinning
- Predictive maintenance
Use cases:
- Process optimization
- Energy & waste reduction
- Quality & yield improvement
Control
Intelligent automation means closing the sensory-motor loop in a way that goes beyond traditional control engineering. In order to fully unlock the hidden potential for increased productivity, controllers must be learned instead of programmed. This means applying Deep Reinforcement Learning to adapt neural network controllers through safe and efficient interaction with a learned process model.
Use cases:
- Process optimization
- Energy & waste reduction
- Quality & yield improvement
