Machine learning
Real-time ML in control technology
Beckhoff offers a solution for machine learning (ML) integrated in TwinCAT 3.
The advantages of system openness through the use of established standards are also available for ML applications. The inference, i.e. the execution of a trained ML model, is possible as a TwinCAT TcCOM object directly in real time, namely for small networks with a system response time of less than 100 µs. The models can be called up via the PLC, C/C++-TcCOM interfaces or via a cyclic task.
Integration into the control technology means that TwinCAT's multicore support is also available for ML. This means that the respective TwinCAT 3 Inference Engine can be accessed from different task contexts. Furthermore, access to all fieldbus interfaces and data in TwinCAT is provided.
You might also be interested in
Analyse und Erfassung von Daten
Datenerfassung in heterogener Umgebung
In Zeiten von I4.0 ist eine durchgängige Datenerfassung in heterogenen Systemumgebungen einer Maschine von großer Bedeutung. Beckhoff hat ein System entwickelt, das den Anwender dabei unterstützt: TwinCAT Scope.
read more...Schwebende Mover mit sechs Freiheitsgraden
XPlanar von Beckhoff ermöglicht neue Wege in der Anlagenkonstruktion. Grundlage sind über beliebig angeordneten Planarkacheln frei schwebende Planarmover. Damit lässt sich eine flexible, exakte und hochdynamische Positionierung erreichen.
read more...Imitation learning in robotics
Flexible automation through human demonstration
Imitation learning makes possible what classic robotics has not been able to achieve so far: fast learning of new processes even without robotics expertise and improved robustness in the face of process variants.
read more...ZDIN: VisionAdapt transfer project
AI recognizes critical chips
In the VisionAdapt transfer project, scientists are researching how camera systems and machine learning can make machining production more efficient, safer and more autonomous. The aim is to automatically detect critical chips and adapt machining...
read more...Innovative method in quality assurance
Federated learning has established itself as an innovative method in the field of quality assurance, especially in scenarios where companies want to protect sensitive production data but still want to benefit from collaborative models.
read more...Reliable robots, cobots and AGVs with embedded AI
Assuming an average machine hourly rate of 2,500 euros, the failure of an industrial robot is associated with high costs. For the first time, local AI makes it possible to reliably implement predictive maintenance from the gearbox to the gripper...
read more...AI vision brings additional opportunities
The seamless integration of AI into the consumer sector has helped to improve our quality of life and make interactions with technology more efficient. In industry, AI vision with machine learning opens up new possibilities that are not covered by...
read more...Classification of component poses
Machine learning in feeding technology
The Institute of Assembly Technology at Leibniz Universität Hannover is developing a machine learning framework for classifying component poses in an aerodynamic feeding system.
read more...AI trends for engineers: where is the journey heading?
According to Gartner, companies that have adopted AI engineering practices have a competitive advantage: by 2026, they could outperform their competitors by around 25 percent in terms of the number and time required to operationalize AI models....
read more...