EI2N 2024 Abstracts

Area 1 - 18th International Workshop on Enterprise Integration, Interoperability and Networking
Full Papers Short Papers

Area 1 - 18th International Workshop on Enterprise Integration, Interoperability and Networking

Full Papers
Paper Nr: 7
Title:

Integrating Knowledge and Data-Driven Artificial Intelligence for Decisional Enterprise Interoperability
Authors:

Christos Emmanouilidis, Sabine Waschull and Jessica Zotteli
Abstract: Although data-driven artificial intelligence (AI) is increasingly applied in decision-making, challenges such as a lack of explainability and trust limit its integration in enterprise decision-making processes. Establishing a minimum level of common information sharing and trust among decision-making stakeholders, often termed as decisional interoperability, is needed for industry adoption. Purely data-driven approaches risk ignoring the enterprise environment and situational context of decisions and are insufficient for such interoperability to the extent that the decision-making rationale is opaque. While linked data and knowledge ap-proaches have long been pursued in the context of data-driven machine learning, these have not been particularly well explored in the context of decisional enter-prise interoperability. This paper aims to narrow this gap. It explores how the introduction of AI is changing decisional interoperability concerns. It then outlines patterns of human-AI teaming in decision-making, as well as methods, such as knowledge-infused AI and mechanisms, such as active learning, for enhancing decisional interoperability. Three diverse application domain-cases offer context for an analysis of AI decision-making considerations and decisional interoperability. The paper concludes with arguments about how integration of knowledge into data-driven AI contributes to decisional interoperability and further work needed in this area.
Paper Nr: 8
Title:

Cognition in Digital Twins for Cyber-Physical Systems and Humans: Where and Why?
Authors:

Yannick Naudet, Jana Al Haj Ali, Ben Gaffinet and Hervé Panetto
Abstract: In modern Cyber-Physical Enterprises, the place of cognition to build more flexible, autonomous, self-adaptive and ultimately more intelligent systems remains unclear. However it could be an essential element to ensure interoperability, as highlighted in recent works on cognitive interoperability. Assuming an environment where cyber-physical systems and humans are coupled with their individual digital twins, we discuss where cognition can exist in the couple <EntityDigital Twin> and how Cognitive Digital Twin and Cyber-Physical Systems can emerge. This is supported by a proposition for a high-level architecture for EntityDigital Twin coupling. The place of Cognitive Architectures as a special model of cognition is discussed. Exploring the integration with the ISO23247 Digital Twin framework for manufacturing, we also highlight its lacks to include cognition and humans.
Paper Nr: 10
Title:

Challenges in Composite Digital Twin Models and their Impact on Interoperability
Authors:

Umar Memon, Wolfgang Mayer, Markus Stumptner and Matt Selway
Abstract: The rapid growth in the domain of Digital Twins has resulted in increasing necessity for the composition of multiple Digital Twin models to represent complex systems. However, this composition poses significant challenges not adequately addressed by current methodologies. Interoperability is one of the primary concerns that occurs during the process of composition. This position paper critically explores current approaches to DT model composition and its correlation with interoperability, identifies the key challenges in the domain, and discusses what further research is required in each challenge. It examines the gaps between currently available solutions and requirements and considers potential approaches that may be extended to address the gap. This paper is believed to further advance the understanding and implementation of composite DT models for complex systems.
Paper Nr: 11
Title:

Research on the Construction Method of Production Equipment Operation Management and Control Information Model Based on Knowledge Graph
Authors:

Jun Li, Keqin Dou, Jinsong Liu, Shunyi Xu, Qing Li and Yong Zhou
Abstract: To address the issues of multi-source heterogeneous information, numerous relationships, and complex interaction logic in production equipment operation management and control(PEOMC) activities. This study proposes a combined approach of "forward engineering + reverse engineering" to construct a general information reference model for PEOMC based on knowledge graph. The forward engineering involves analyzing the information resources related to PEOMC functions to design an information meta-model. The reverse engineering involves mining, analyzing, and refining multi-source and heterogeneous PEOMC engineering practice data. Guided by the information meta-model, a knowledge graph is constructed. By mapping and transforming entities, attributes, and relationships from the knowledge graph to the information model, a general information reference model for PEOMC is established. Finally, taking the information model construction of the predictive maintenance scenario in an aero-engine transmission unit manufacturing workshop as an example, the effectiveness and scientific validity of the method proposed in this study are verified.
Short Papers
Paper Nr: 5
Title:

Human-Centric Smart Manufacturing: A Framework Utilizing Digital Twins and Cyber-Physical Production Systems (CPPS)
Authors:

Marcelo Feliciano Filho, Anderson Luis Szejka and Eduardo Rocha Loures
Abstract: The shift to Industry 5.0 emphasizes human-machine collaboration, personalization, sustainability, and worker support. This paper explores integrating Digital Twins (DT) and Cyber-Physical Production Systems (CPPS) within a human-centric Smart Manufacturing framework. Digital Twins offer real-time prediction, simulation, and optimization, enhancing production efficiency and adaptability. CPPS merges computational and physical processes, supporting human-machine collaboration and improved decision-making. The main objective is to enhance sustainability, efficiency, and human-centricity in data-driven manufacturing. This research reviews current technologies, proposes a novel framework integrating DT and CPPS, and assesses implementing a Smart Product (SP) within a Smart Factory (SF) scenario. Key innovations include designing and integrating SPs, developing smart manufacturing processes, and applying DT for enhanced SP functionality. The findings demonstrate improvements in production efficiency, customization, human-machine collaboration, and sustainability, aligning with Industry 5.0 principles. The study concludes with recommendations for future research to integrate ergonomics, cybersecurity, and ethical considerations in smart manufacturing.