Generative AI models such as language models based on the transformer architecture have shown the capability to generate text, specifications and source code. Using techniques such as prompt engineering and retrieval-augmented generation, these models can be equipped with domain knowledge and can be used to generate domain-specific output. The OPC UA specifications and OPC UA information models contain domain knowledge and the goal of this working group is to develop best practices, guidelines, examples and data files to simplify the use of Generative AI to support the use of OPC UA.
The working group will publish a guidelines and examples document for the purpose of exposing domain-specific and general information based on OPC UA constructs and specifications as inputs for generative AI models. The working group does not intend to produce domain-specific Generative AI models for commercial use. It may provide examples of and instructions for training such models.
Initially, the working group will analyze the following use cases for OPC UA for AI
- The use of Generative AI during the engineering phase of systems based on OPC UA and its companion specs
- The use of Generative AI as a tool for analyzing data originating from OPC UA systems and based on companion specs
- The use of Generative AI as part of an user interface of control devices based on OPC UA and its companion specs
go back