Manufacturing Systems, Processes and Enterprise Interoperability
Market globalization and economic turmoil have made more than clear the need to increase the competitiveness of the European and Greek industry and enterprise. Technological advances such as the ubiquity of networking and computational infrastructure create new opportunities and challenges. The evolution of the Industrial Internet of Things (IIoT), as the application of the Internet of Things (IoT) in the industrial domain is called, creates a new environment for the industry to operate in.
Dealing with the IIoT evolution is a primary axis of the research team effort. The number of IoT connected devices worldwide is expected to grow from 15 billion in 2015 to 31 billion in 2020 and 75 billion in 2025, while the overall IoT market is expected to grow by 1 billion USD annually from 2017 onwards. IIoT faces additional integration and interoperability challenges with reference to the IoT, as it needs to efficiently integrate Operational Technology, represented by PLCs, NCs, robot controllers and SCADA, with Information Technology, having followed different evolutionary steps.
Vertical integration and interoperability challenges of the different systems / applications that need to cooperate in the industrial manufacturing environment from the ERP level down to the field level remain valid, contributing to the overall manufacturing environment flexibility and agility. Model driven interoperability, unified modeling approach for products, processes and resources, and IIoT integration represent major goals. Seamless integration and interoperability among enterprises enables new opportunities and advanced business models such as collaborative manufacturing, integrated supply chain management, smart specialization. Scientific issues associated include (i) enterprise modeling for enterprise interoperability, comprising requirement engineering, service modeling for businesses, metamodeling, model synchronization, (ii) semantics for enterprise interoperability comprising enterprise application analysis and semantic elicitation, reasoning, methods and tools for model transformation and data reconciliation, semantic mediation and enrichment of enterprise models, semantic web-based approaches, (iii) architectures and frameworks for interoperability comprising agent-based approaches, enterprise application integration, model driven architectures, service oriented enterprise architectures, ubiquity, mobility and large scale interoperability.
Industrial Control Systems represent a sub class of Cyber Physical Systems (CPS) primarily addressing issues in the industrial manufacturing domain. In this context, general challenges of CPS are applicable: predictability, reliability, sustainability, dependability, security, and interoperability. The research team focuses primarily on Industrial Control Systems interoperability aspects and their attributes: composability - the ease of incorporating operating components, scalability - the ease of scaling in size and throughput, and heterogeneity – the ease of combining different components. Service Oriented Architecture application at all levels of the industrial manufacturing environment is sought, utilizing standardized interfaces such as OPC UA at device level.
Industrial Control Systems are not dedicated to the industrial manufacturing domain, but are also spread to other domains that present similar challenges at their field level. Critical infrastructures are such a domain, comprising diverse application areas such as Smart Buildings, Smart Energy, Transportation Infrastructure Safety and Security, water management, etc. The research team focuses especially on the Smart Cities paradigm and more specifically on the Smart Energy and Smart Building domains and their integration, interoperability and data analysis challenges. Open data and APIs for the integration of Smart Cities related pilot infrastructures contribute to the Smart City paradigm enhancement. In the context of Smart Buildings research effort is placed towards the Living Lab concept, a user-centered, open innovation ecosystem integrating research and innovation processes. Embedded systems, networks, IoT, different computing paradigms and Big Data analytics are necessary components.
IIoT importance has led to the emergence of different initiatives offering relevant reference architectures to facilitate its application primarily in the industrial manufacturing domain and secondarily in other domains that present similarities. Such initiatives include Platform Industry 4.0, the Industrial Internet Consortium, and Society 5.0. The research team is aligning its activities towards these efforts.
The Cyber Physical System paradigm and the Internet of Things has transformed computational services to a commodity. The blurring boundaries between the different types of nodes due to their embedded intelligence has led to different computing paradigms offering different characteristics: cloud, fog, edge and mist computing. Real timeliness, quality of service, latency and mobility of nodes are some of the issues that determine the selection of computing model or mix of models. Artificial Intelligence (AI) applications are expected to promote computing paradigms closer to the end nodes, especially in the case of applications with strict timing requirements for big data analytics, such as for instance deep learning.
The Digital Twin, one of the enabling technologies in the context of Industry 4.0 and IIoT, represents a digital model accurately representing its physical counterpart, evolving and continuously updating to reflect changes in the physical world. It offers a holistic approach for dealing with different issues in the manufacturing environment ranging from simulation of real world scenarios, to lifecycle issues, to security issues.
Accomplishments of the research area include:
- Mechatronical device structure and Integrated Distributed Environment supporting dynamic distribution of control application code (based on IEC 61499) and management functionalities at the shop-floor level
- Integrated production process, product and resource modeling for the manufacturing environment and relevant manufacturing environment ontology and description language (P5DL)
- Multi agent system for increasing the autonomy in the manufacturing environment implementing the GAIA model at the Manufacturing Execution System (MES) layer
- Extension of the GAIA model in the framework of the Sensing Enterprise towards integration of the IIoT
- Framework for adaptive multi agent systems for manufacturing
- Cloud Manufacturing platform exposing manufacturing resources as services facilitating collaborative manufacturing
- Manufacturing system interoperability demonstrator based on a serious gaming platform, enabling Hardware-in-the-Loop simulations and utilized in the framework of scenarios like asset management and preventive maintenance
- Framework for enterprise process implementation in terms of workflows, enterprise ontologies and service oriented architecture
- Collaborative Continuous Replenishment Model based on the RosettaNet Framework
- Methodological framework for the support of Virtual Enterprises
- Semantically enriched platform for the support of the tourist sector viewed as a virtual cluster of thematically clustered enterprises
- Integrated Supply Chain Management solutions for the provision of end-to-end safety and security solutions in transport and logistics operations
- Matlab / Octave calculation-based quasi-steady-state model for building energy consumption estimation