Título: Seamless Integration of Assets into Industry 4.0 Intelligent Systems Data and Model Interoperability Analysis
Autores: Vitor Furlan de Oliveira, Ogobuchi Daniel Okey, Gabriel Venturini, Gabriel de Oliveira Pereira, Fabrício Junqueira, Claudio José Bordin Júnior & Rômulo Gonçalves Lins
Resumo: The evolution of Industry 4.0 demands seamless integration of intelligent assets capable of interoperating across various systems and contexts. This paper proposes an interoperability evaluation in intelligent systems structured according to the Industry 4.0 principles, namely the Reference Architecture Model Industrie 4.0 and the Asset Administration Shell as a digital representation standard. Two intelligent system architectures were developed and compared: a centralized approach emphasizing data interoperability and a federated learning setup demonstrating model interoperability. Both scenarios were validated using a realistic predictive maintenance dataset, and interoperability was assessed through metrics related to model performance, computational cost, and data quality. Semantic interoperability was also qualitatively analyzed, highlighting the importance of standardized data dictionaries. The results show that while the centralized model achieved higher predictive performance, the federated approach provided computational efficiency and data privacy without loss of interoperability. The study also highlights the impact of technology-related constraints, such as limitations of the machine learning framework, on the outcomes of interoperability. These findings demonstrate the viability of AAS-based architectures for developing intelligent, interoperable Industry 4.0 systems.
Palavras-chave: Interoperability; Industry 4.0; Intelligent Systems; Federated Learning.
Páginas: 8
Código DOI: 10.21528/CBIC2025-1191417
Artigo em PDF: CBIC_2025_paper1191417.pdf
Arquivo BibTeX:
CBIC_2025_1191417.bib