FabOS component
Component identification
Component identification solution that uses a neural network and vector embedding approach to quickly and accurately identify components in real time.
Project partners involved
Component identification
Unlike other solutions, new components no longer have to be trained and integrated at great expense, but can be called up directly.
Project partners involved
Problem:
- Fast, dynamic changes in production
- Time intensive training of the AI
- Data protection and operational security
- Real-time detection
Solution approach:
- Neural network trains AI on content-based similarity matching
- Network produces comparable and similar vectors for the same part
Functions
Quality control of individually manufactured parts
Identification and assignment to the blueprint in real time
No separate training for new components required
Quality control of individually manufactured parts
Identification and assignment to the blueprint in real time
No separate training for new components required
Use cases
FabOS forms the foundation of a modern production environment
Bin Picking
- Automatic separation of chaotically filled boxes
- High performance for uninterrupted line production
- Automatic identification of unknown parts
- Adaptation to different hardware for commissioning
FabOS components used:
Bin Picking
Project partners involved:
Your contact for questions
Grischa Hauser
COMPAILE Solutions GmbH
https://usebasin.com/f/08662176d5f8
Discover more FabOS components
Secure On - & Offboarding
Transfer App
Speech recognition
Convergent network
One Stop Shop
Seamless Computing
Service Lifecycle Management
Asset Administration Shell