Beckhoff has recently developed a TwinCAT Chat client for the TwinCAT XAE development environment. This allows for convenient integration of large language models (LLMs) such as OpenAI's ChatGPT into the TwinCAT project development process, aiming to enhance the efficiency of control program programming. The client also offers optimization potential by directly supporting functionalities.
Large Language Models (LLMs) are neural networks trained on vast amounts of text. In recent years, their applications in natural language processing have become widespread, encompassing chatbots and machine translation.
The TwinCAT Chat client automates programming tasks in an AI-supported manner, including the creation or addition of function block code. It can even refactor and optimize existing code and generate documentation. Implemented in TwinCAT XAE, the client connects to various LLM clouds (e.g., ChatGPT based on Microsoft's AzureTM), providing a user interface. It communicates with the PLC development environment through the Automation Interface, accessible via a dedicated chat window in Visual Studio. The client is specifically optimized for TwinCAT 3 users, offering extensive support for LLM functionality tailored to TwinCAT's specific context.