Due to the current health situation related to COVID-19 and the restrictions imposed to Higher Education activities, the SOHOMA 2022 Workshop will be held entirely online on 22-23 September 2022, with the detailed program displayed in the menu PROGRAM of the SOHOMA 2022 website.
The hours for paper presentation correspond to the Romanian time. The technical sessions will be organized in MS Teams. The information about the access to the workshop’s sessions will be provided in due time.
The fees are reduced to those for online participation.
About the Workshop
It is our pleasure to invite you to participate in the 12th International Workshop on Service Oriented, Holonic and Multi-Agent Manufacturing Systems for Industry of the Future – SOHOMA’22, organized by University Politehnica of Bucharest (the Faculty of Automatic Control and Computers and the Research Centre in Robotics and CIM – CIMR) in collaboration with Polytechnic University Hauts-de-France (the LAMIH Laboratory of Industrial and Human Automation Control, Mechanical Engineering and Computer Science) and Polytechnic Institute of Bragança (the CeDRI Research Centre in Digitalization and Intelligent Robotics). The SOHOMA’22 workshop has the scientific support from the IEEE-IES Technical Committee on Industrial Agents, the GDR MACS-IMS2 Research Group on Intelligent Manufacturing Systems and Services and from the General Association of Engineers in Romania AGIR.
The main objective of SOHOMA workshops is to foster innovation in smart and sustainable manufacturing and logistics systems and in this context to promote concepts, methods and solutions for the digital transformation of manufacturing through service orientation and agent-based control with distributed intelligence.
SOHOMA 2022 will take place in Bucharest, Romania at University Politehnica of Bucharest, Faculty of Automatic Control and Computers on September 22-23, 2022.
The Workshop’s Theme
The theme of the SOHOMA’22 Workshop is “Virtualisation – a multifaceted key enabler of Industry 4.0 from holonic to cloud manufacturing”.
Virtualization in industrial control and supervision can be viewed as a multi-purpose, multifunction key enabler of the digital transformation addressed by the Industry 4.0 framework for the factory of the future. It refers to the creation of virtual, software versions of computer hardware platforms, computer network resources, industrial equipment, control systems, processes and activities: process planning, scheduling and execution, product routing, making and tracking. In the manufacturing domain, two technologies are used to create virtual instances of physical devices and non-material entities on three layers: shop floor, automation and Manufacturing Execution System: digital twins and the holonic paradigm.
At present, the digital twin concept evolved to a highly advanced modelling and simulation technology used for enhanced product and process design, event-driven resource maintenance and health monitoring, anomaly detection, and intelligent manufacturing control with prediction of behaviours, production costs, reality-awareness and optimization. Modelling reality is done with two classes of digital twins (DT): data-driven (with physical counterpart) that relies on distributed edge computing and storage of shop floor data in Industrial IoT frameworks, and model-driven (without physical twin) – a digital simulation for design and analysis. The multi-purpose character of digital twin virtualization in manufacturing derives from its main application areas:
- Simulation with software in the loop: a model-driven DT with control software operating in simulation exactly as it will be deployed; the virtual model is used to design equipment, tune parameters, validate layouts and compare performances in special conditions.
- Synchronizing the virtual twin with its physical counterpart: this type of DT is involved in all activities that imply its physical twin (e.g., shop floor resource, product with embedded intelligence) and provides control with situation awareness such as health monitoring or diagnostic.
- Embedded simulation faster than real time: a combination of a model-driven DT with physical, non-material twins of activity instances (e.g., individual product execution orders) allowing the latter to virtually execute much faster than real-time (at computer speed) the intentions of the decisions makers (e.g., operation scheduling and allocation to resources, cost optimization, control tuning) on the twins of the resource instances, offering predictive situation awareness.
The distribution of intelligence within industrial control system and the need for collaborative decisions of strongly coupled shop floor devices in manufacturing domain) led to the adoption of a new modelling approach for robust and optimized process control with agent-based implementing framework: the holonic control paradigm. This approach is based on the virtualization of sets of abstract entities: products (reflecting the client’s needs, value propositions), resources (technology, humans – reflecting the producer’s profile and skills) and orders (reflecting business solutions) that are modelled by autonomous holons collaborating in holarchies by means of their information counterparts – intelligent agents – to reach a common, production-related goal.
Virtualisation is a multifaceted technology from the point of view of addressed manufacturing workloads: individual vs. collective, respectively local vs. global workloads:
- The agents that virtualize the generic classes of factory entities: resources, products, orders and processes instantiate them at production run for two types of local workloads – (a) individual resource state, behaviour and performance monitoring, respectively tracking product execution; (b) collaborative activity scheduling and allocating on resources for products that add up to the factory’s capacity of simultaneous execution.
- The set of workloads connected to MES computing entities that are seen as expert or advisor agents acting as optimization engine, machine learning predictor and outlier detector; these are global workloads relative to the full production stage at the farthest batch horizon.
- Virtualization represents the main enabling technology for cloud manufacturing, i.e., virtualizing global computing-intensive, real time MES workloads in classical hypervisor virtual machine approach combined with container-based techniques.
Virtualization of manufacturing devices, products, activities and information systems assures the premises for the smart factory of Industry 4.0 which is based on cyber-physical production systems – the new generation of manufacturing systems conceived to be adaptive, fully connected, analytical and highly efficient.
The research of the SOHOMA scientific community is aligned to these technological developments that build up actual trends and development priorities for cyber-physical systems in the manufacturing, supply chain and logistics industries:
- Sustainable Holonic Architectures (HCAs) able to reduce the energy consumption of manufacturing systems, to optimize the production costs with reality-awareness, adaptability at environment changes and robustness at technical disturbances.
- The factory data streams and global MES functions will be mapped to specific workloads in the cloud, defined in terms of activity scheduling, resource assignment and behaviour forecast; the latter incorporate AI and ML capabilities. The industrial sector is interested in deploying autonomous workloads to achieve higher productivity and better operational safety.
- Developing the four key drivers to sustain the paradigm Logistics 4.0: data automation and transparency (end-to-end visibility over the supply chain, logistics control tower, optimization soft-ware); new production methods (robotized palletizing, stereovision-based picking); new methods of physical transport (driverless vehicles, autonomous pickers, drones); digital platforms (shared warehouse and transport capacity, cross border platforms).
- Manufacturing as a Service (MaaS) – the new models of service-oriented, knowledge-based manufacturing systems, virtualizing and encapsulating shop floor and MES workloads into cloud networked services – will also address product design for ‘open manufacturing’, and the knowledge and infrastructure sharing in cloud collaborative, universal manufacturing enterprises.
- Developing Industry 4.0 models and frameworks using digital industrial technologies: advanced robotics, additive manufacturing, augmented reality, extended digital modelling and simulation through digital twins, horizontal / vertical enterprise integration, cloud, cybersecurity, Big Data and Analytics, and the Industrial Internet.
This 12th SOHOMA edition puts the focus especially on how the core production, digital and cyber technologies are applied, interconnected and orchestrated in order to create a product-service centric closed loop collaboration covering the design, engineering, production, distribution and after-sales phases in the modern manufacturing value chain.
This approach derives from the research performed in the last years by members of the scientific community SOHOMA, based on recently developed key digital and cyber technologies – cloud and fog computing, digital twins, edge computing, optimization, cognitive robotics, machine vision, additive machining, artificial intelligence and machine learning – and ethical and societal models in future industrial systems – human-machine cooperation, human integration in cyber-physical systems, ethics of the artificial, humans in Industry 4.0, low-cost digital solutions for manufacturing:
- Smart technologies and secure connectivity in factories, i.e., the digital integration of manufacturing and logistic equipment in terms of information, communication and automation technologies.
- CPS-enabled reconfiguration of automated manufacturing systems: (1) Deployment of legacy production equipment and systems; (2) Increasing autonomy and intelligence of existing machinery and robots; (3) Adaptation through context awareness and reasoning aiming at making machinery and robots aware of their surroundings; (4) Developing a multi-layered, decentralized control architectures in which resources can take autonomous decisions.
- Intelligent decision making in cloud manufacturing through big data streaming and machine learning; combining data-driven digital twins for predictive situation-awareness with model-driven digital twins simulating the reality of interest faster than real-time with software in the loop.
- Sharing of data/information from all the supply chain’s elements to support continuous monitoring and automatic control of all the production phases while preserving security and confidentiality of data shared along the supply chain.
- Data mining and real time analysis to design new product-service systems, and redesigning business models for ecosystems of product-service starting from data collected during the product life cycle.
- The adoption of IoT and CPS as enablers of the digital transformation of processes in the manufacturing value chain.
- Applied ethics in industrial cyber-physical systems; ethics-aware design, engineering, control and supervision activities.
- Human integration in modern manufacturing environments; automated administrative logistics processes for human integration in industrial cyber-physical systems.
- Service Manufacturing which includes design for open manufacturing, optimization, maintenance, supply and distribution activities, all of them being offered in the “as a Service” option. service manufacturing was proposed.
- Fostering the open manufacturing enterprise – responsive to the X-as-a Service model, where X covers design, manufacturing, supply, and distribution, and supports resource sharing and networking.
Papers presenting solutions based on these technologies for new applications in the manufacturing value chain at the confluence of information technology and automation are especially welcome. All contributions must indicate alignment with the workshop’s theme.
- Modelling of discrete event dynamic industrial systems
- Multi-Agent Systems and control with distributed intelligence in industry
- Product intelligence: concepts, architectures, implementation, use cases
- Holonic Manufacturing Execution Systems
- Intelligent Manufacturing Systems
- Dynamic and green infrastructure for sustainable manufacturing
- Mixed production planning and scheduling
- Virtual factory, networked supply chains and customer-oriented logistics
- Manufacturing Integration Framework
- Digital Twins with resource and product virtualization
- Edge and Fog Computing for Industrial Internet of Things
- Industrial Internet of Things / Physical Internet
- Cloud Manufacturing and resource virtualization
- Servitization and Product-Service Systems
- Swarm intelligence in manufacturing
- Multi-robot systems in manufacturing, control and applications
- Bio-inspired theories for smart manufacturing and evolutionary robotics
- Big Data streaming and the contextual enterprise
- Artificial intelligence and machine learning in large scale manufacturing
- Predictive resource health monitoring and maintenance
- Cyber-Physical Production Systems and Industry 4.0
- Manufacturing as a Service, product design for open manufacturing
- Universal Manufacturing: models, cloud manufacturing, enterprise specifications
- Digitalization of supply chains and Logistics 4.0
- Reconfigurable manufacturing systems
- Humans-Systems integration in Industry 4.0 environments
- Ethics of autonomous intelligent systems
Publication of Workshop papers
All papers accepted for presentation will be included in the SOHOMA’22 proceedings volume and published in the Springer featured book series “Studies in Computational Intelligence”.
The Proceedings volumes of SOHOMA’11-SOHOMA’21 previous events, published in special issues of the Springer Book series “Studies in Computational Intelligence” have been included in Web of Science, Scopus, ISI Proceedings and the DBLP Computer Science Bibliography (University of Trier, Germany).