Data Mapping Framework for the AAS

Seamless interoperability in industrial systems

Abstract

Achieving seamless interoperability among heterogeneous industrial systems remains a fundamental challenge in the deployment of Industry 4.0 environments. Conventional integration methodologies impose significant configuration overhead and demonstrate limited adaptability to diverse data structures. This paper proposes an automated semantic data mapping framework targeting the Asset Administration Shell (AAS) metamodel, wherein large language models (LLMs) are employed under a zero-shot learning paradigm to generate integration rules at inference time.