New software addition enables machine learning integration
SOFTWARE & SYSTEMS BECKHOFF LAUNCHES TF38
BECKHOFF
New software addition enables machine learning integration
Automation and control supplier , Beckhoff , has launched TF38 , a new addition to its TwinCAT 3 automation software package . TF38 seamlessly integrates machine learning ( ML ) capabilities into industrial automation processes , setting a new standard for intelligent manufacturing .
The TF38 function offers a highperformance execution module , or inference engine , designed for the real-time application of trained machine learning models . By supporting established frameworks such as SciKit-Learn , libSVM , and XGBoost , TF38 enables the execution of a wide array of ML algorithms directly within the automation environment . This integration allows for enhanced data analysis , predictive maintenance and optimised production processes .
A key feature of TF38 is its support for the open neural network exchange ( ONNX ) format , an open standard that facilitates interoperability between various ML frameworks . This ensures that models trained in diverse environments can be seamlessly deployed in TwinCAT 3 , providing flexibility and efficiency in developing and implementing ML solutions .
“ The introduction of TF38 is a significant milestone in the convergence of machine learning and industrial automation ,” explained Beth Ragdale , product manager at Beckhoff UK . “ Embedding advanced analytical functionalities directly into our control systems enables our
The introduction of TF38 is a significant milestone in the convergence of machine learning and industrial automation .
customers to achieve exceptional levels of performance and efficiency .”
The TF38 function is designed to process all inputs , outputs and internal data from the automation controller using machine learning algorithms . This capability not only improves product quality but also supports real-time classification of machine operational states . By training models with historical data , the system can execute ML models on PC-based controllers , allowing immediate actions to be implemented within the automation system .
Beckhoff ’ s innovation is further demonstrated by the integration of TF38 into the existing TwinCAT 3 environment . Users can leverage familiar tools and interfaces to develop , train and deploy ML models , reducing the learning curve and accelerating the adoption of advanced analytics in industrial applications .
For more information about TF38 and how it can transform your automation processes , visit the Beckhoff website .
For further information , please visit www . beckhoff . co . uk
84 PECM Issue 73