Periodic Reporting for period 1 - COMPOSITION (Ecosystem for Collaborative Manufacturing Processes ? Intra- and Interfactory Integration and Automation)

Summary
Requirements of modern production processes stress the need of greater agility and flexibility leading to faster production cycles, increased productivity, less waste and more sustainable production. At the factory level, decisions need to be supported by detailed knowledge...

Summary

Requirements of modern production processes stress the need of greater agility and flexibility leading to faster production cycles, increased productivity, less waste and more sustainable production. At the factory level, decisions need to be supported by detailed knowledge about the production process and its interplay with external entities. Unfortunately, historical and live data that generates this knowledge is becoming more and more distributed and few solutions are available that can easily tackle the implied challenges. Moreover, factories are becoming less isolated in the productive tissue of nations and several suppliers and third party service providers need to be contacted and coordinated to implement decisions taken at the factory level.
In such a worldwide and dynamic environment, the ability of automatizing the preliminary coordination and negotiation activities involved in setting up supply chains for specific needs, in an open marketplace-like fashion, could greatly improve the ability of factories to quickly react to external challenges and driving forces.
COMPOSITION has two main goals: The first goal is to integrate data along the value chain inside a factory into one integrated information management system (IIMS) combining physical world, simulation, planning and forecasting data to enhance re-configurability, scalability and optimisation of resources and processes inside the factory. The second goal is to create a (semi-)automatic ecosystem, which extends the local IIMS concept to a holistic and collaborative system incorporating and inter-linking both the Supply and the Value Chains.

Work Performed

WP1
Drove activities based on recommendations and outcomes of the 1st review. Drove developments towards use cases
Kept organizational frame and collaboration tools running.(most important: monthly status telephone conferences, management of high level tasks in Jira)
2 online and 3 physical meetings and 2nd review organized
Established procedure to assure interrelations between pieces of work are more clearly described
Project progress monitoring and quality check: 28 deliverables and 8 milestones with no or slight delay
WP2
Use Cases revisited, revised and prioritised to support the developmental efforts
COMPOSITION Architecture developed and revised in line with the evolution of the solutions and components taking place in the technical WPs
Lessons Learned posted in the Repository and the associated changes to the requirements in the JIRA database reviewed
WP3
A common methodology and respective notations for modelling processes and stakeholders have been established and process models of the industrial processes defined using BPMN
Digital Factory Models (DFMs), which aim at the digitalization of industrial aspects developed. Structure of DFM schema based on popular standards
Interaction of the DSS with the simulation and prediction engine defined
A dockerized demo version was implemented. This version included an example of the rule engine performance which is based on Finite State Machines
A first visualization process of the rule engine results took place
WP4
Deployment and initial configuration of two instances of Authentication Service (Keycloak) and Nginx reverse-proxy at the docker server
Development and initial testing of RAAS service. Initial deployments of prototypes to validate the development
Modifications on Authorization Service (EPICA) to integrate with Authentication Service
Development of cybersecurity agent based on neural networks
Propose the use of Multichain blockchain to store and share the hashes of messages and the public keys to be used to validate the signature of messages
Definition of reputation model and use Multichain blockchain for storing and sharing stakeholderĀ“s ratings
WP5
Deliver Learning Agents Framework, Deep Learning Toolkit and Intra-Factory Interoperability Layer
Kept up with the support role of HMIs design and sensors recommendations
WP6
RabbitMQ extended with integrated security framework and blockchain-based log functionality
Extension for RabbitMQ support for request-response messaging using REST
First version of the Matchmaker and marketplace ontology developed
Agent on stakeholder side and connection to the marketplace have been developed
WP7
Test, installation and operation plan template was shared with other FoF11 project groups
WP8
An evaluation framework developed measurable metrics and guidelines for validation
Initial tests of sensors at KLEEMANN?s and BSL?s shopfloors successfully completed
WP9
Attended 27 External workshops, seminars, etc. in total, 5 of them in symbiosis with other FoF-11 projects and EFFRA
Organised 37 online meetings and 1 physical meeting
Issued two newsletters and preparing another one
6 publications submitted

Results

Collaborative manufacturing ecosystem
-Clustering of brokers behind a load-balancing system
-Broker as dynamically scalable cloud service
(Agent-based) marketplace
-Language-agnostic agent implementation
-Automatic bidding process using matchmaking techniques
-Semi-automatic formation of new supply chain using agents
Multi-level modelling and simulation
-Addressing the problem of developing a holistic and complete data model for representing manufacturing systems: multi-scaled Digital Factory Models
-Mapping between factory resources and the corresponding factory processes
-Enables the creation of a factory monitoring framework based on BPMN modelling of processes
-Complete and holistic solution for malfunctions prevention
Blockchain, trust, traceability
-Configurable for consortium or private use
-Speed up the generation of new blocks
-Horizontal or vertical partitioning of the log
Multi-level Big Data Analytics
-Applying continuous learning on Long Short-Term Memory artificial neural networks, in an untrained manner, in contexts where no historical data have been made available by end users
-Technologies and algorithms for on-the-fly, online, real-time, storage-less data analysis for discrete sensor data streams
Secure and Trustworthy Framework
-RAAS service to override RabbitMQ message broker built-in authentication and authorization mechanisms, and use Security Framework services
-Use blockchain technology to store stakeholder?s ratings, based on the reputation model defined, to implement a simple PKI (Public Key Infrastructure) and to log messages, by storing messages hashes and metadata; and be able to validate the content
Attached files
No attachments available
Ecosystem for Collaborative Manufacturing Processes _ Intra- and Interfactory Integration and Automation
Contacts
Not specified (see website if available)
Organisations
  • Not specified