Connecting AAS and LLMs for Controlling NDT Devices

Full abstract

Read the full abstract

The ongoing development and implementation of artificial intelligence (AI) is opening up new ways to increase efficiency and improve quality of processes, including the process of non-destructive testing (NDT). Traditional AI applications, which focus on analyzing existing processes, are recently being supplemented by generative AI models that dynamically generate content such as text, images, or videos and thereby facilitate intuitive human-machine interactions. The following report describes how generative AI is used in non-destructive testing (NDT) to optimize testing processes through more efficient data processing, precise information provision, and bidirectional AI-machine interaction. This includes the development of an assistance system that processes NDT-relevant expert knowledge and integrates device information to optimize inspection processes and support experts. The use of technologies such as the “asset administration shell” (AAS) and “OPC UA” (Open Platform Communication Unified Architecture) enables real-time data manipulation and admits bidirectional interaction with the testing devices. The dynamic adaptation of the system to user needs allows testing devices to actively intervene in the testing process, or, with regard to digital certificates, to check the suitability of the device or create digital test reports.