The Products and Components developed in COMPOSITION are listed below. Each of the 19 entries contains a brief description and contact details for more information. To see where the individual component belongs in the COMPOSITION Solutions, you can download deliverable D2.4 The COMPOSITION Architecture Specification II.

Agent Marketplace Framework

The COMPOSITION Marketplace is a fully distributed multi-agent system designed to support Industry 4.0 exchanges between stakeholders. It is particularly aimed at supporting automatic supply chain formation and negotiation of goods/data exchanges, granting trust and security in every negotiation step by automated agents.

Contact: Claudio Pastrone, LINKS Foundation
claudio.pastrone@linksfoundation.com

Anti-Forge Stamp

The Anti-Forge Stamp developed for the COMPOSITION platform is based on blockchain (Multichain). It is a distributed registry aimed at avoiding tampering with audit operatives. Many protocols, e.g.,  KNX, ModBus and MQTT, are supported as are publish/subscribe and request/response messaging.

Contact: Rodrigo Díaz, Atos Spain SA
rodrigo.diaz@atos.net

Asset Tracking System

The Asset and Component Tracking System is used for tracking of lost assets or components, e.g., lost or misplaced equipment due for pre-use calibration.

Contact: Peter Haigh, Tyndall National Institute
peter.haigh@tyndall.ie

Building Management System and BMS Device Connectors

The Building Management System (BMS) is a complete and fully qualified stand-alone product for a single installation. In COMPOSITION, the BMS provides components for building local smart environments consisting of a number of devices, applications and services, which can be discovered and communicated with using the publish/subscribe or request/response messaging. The BMS device connectors provide integration of heterogenous devices in the middleware, implementing the functionality of the Device Integration Layer. Due to the diversity of available IoT devices and possible integration scenarios, it is a concept rather than a single component.

Contact: Matteo Pardi, Nextworks
m.pardi@nextworks.it

Collaborative Manufacturing Services Ontology and Language

With the Collaborative Manufacturing Services Ontology and Language, an enterprise can model and describe both manufacturing concepts (services, tools, materials, resources, machines, etc.) and supply chain concepts (price, delivery time, payment method, etc.) It supports flexible specification and execution of manufacturing collaboration schemes and describes relations between business entities. It can be used as the knowledge base for developing matchmakers that match offers and requests for specific domains in manufacturing.

Contact: Dimosthenis Ioannidis, CERTH ITI
djoannid@iti.gr

COMPOSITION Agents

The COMPOSITION Agents are autonomous software entities able to operate on the COMPOSITION Agent Marketplace. They can be tailored to the specific stakeholder needs and support several negotiation protocols/algorithms, allowing stakeholders to choose what is best for their needs.

Contact: Claudio Pastrone, LINKS Foundation
claudio.pastrone@linksfoundation.com

Decision Support System

The Decision Support System (DSS) for Optimising Manufacturing Processes receives events from various registered sources (sensors, forecasting, simulation tools). It is a simple plugin system that suggests optimal solutions. Based on Finite State Machine methodology, it uses a rule engine to dynamically assess available data and provide suggestions, based on business priorities and objectives. The DSS supports decentralised decision making and improves overall system efficiency.

Contact: Cosmas Vamvalis, Atlantis Engineering SA
vamvalis@abe.gr

Deep Learning Toolkit

The Deep Learning Toolkit (DLT) is an ad-hoc solution tailored to the user’s specific needs. The DLT utilises machine learning and deep learning methodologies for multiple-level analysis on heterogeneous sources. It provides custom support to intelligent decision and evaluation systems.

Contact: Luigi Giugliano, LINKS Foundation
luigi.giugliano@linksfoundation.com

Digital Factory Models

The COMPOSITION Digital Factory Model (DFM) is a web-based component, incorporating well-known standards (B2MML, gbXML, OGC, x3d, BPMN). The DFM enables modelling of all industrial aspects, such as definition of factory processes and factory resources (assets, equipment, sensors, buildings, actors, etc.) Through a web service API, data exchange between analytic tools and decision support systems is facilitated by the high level of simplicity, extensibility, interoperability and openness.

Contact: Dimosthenis Ioannidis, CERTH ITI
djoannid@iti.gr

Ecosystem Blockchain

The deployed blockchain implementation has been adapted as the baseline of its log-oriented architecture. The blockchain provides an audit trail for manufacturing and supply chain data, with controlled access to the supply chain. It is based on a consortium blockchain and is not dependent on a trusted central authority. It is an open system, with a shared ledger of business transactions.

Contact: Peter Rosengren, CNet Svenska AB
peter.rosengren@cnet.se

EPICA

EPICA is a multi-stakeholder attribute-based access control and authorisation engine using the XACML standards. This engine was released as open source with Apache 2.0 licence. EPICA is applied to secure messaging for IoT devices and services by RabbitMQ, with authentication provided by integration with open source keycloak.

Contact: Rodrigo Díaz, Atos Spain SA
rodrigo.diaz@atos.net

Fan Conditional Monitoring System

The Fan Conditioning Monitoring System enables predictive maintenance of machines on the factory floor to minimise cost and machine downtime through better prediction of failures or malfunctions. The system assists in deciding the optimal time for performing maintenance activities and allocation of resources.

Contact: Peter Haigh, Tyndall National Institute
peter.haigh@tyndall.ie

Fill-Level Sensors

Deployment of solutions for monitoring the fill level of bins with recyclable material trigger handling of the material and the involvement of the recycling companies. The sensors trigger the online bidding process in the Virtual Marketplace.

Features include:
Easy installation with a LoRa Gateway.
Durable case installed in industrial bins.
Error checking for false measurements and overload notification capabilities
Provision of data to Simulation and Prediction tools.
Use of ultrasonic sensor modules.
Battery powered with ultra-low power design

Contact: Dimosthenis Ioannidis, CERTH ITI
djoannid@iti.gr

LinkSmart® Learning Service

The LinkSmart® Learning Service is an Industry 4.0 agent for stream mining and machine learning in real time. It provides mining through big data streams and composed contextualized events and enables various kinds of in-memory data processing, ranging from data annotation and aggregation to complex event processing. It is open source and suitable for high throughput processing of manufacturing data.

Contact: J. Angel Carvaja Soto, Fraunhofer FIT
jose.angel.carvajal.soto@fit.fraunhofer.de

Private Blockchain

The Private Blockchain implementation provides an audit trail for manufacturing, based on private blockchain, using the blockchain ledger to track all movements and handling transactions of a component or asset inside the factory. It can be integrated with the Real-time location tracking system.

Contact: Peter Rosengren, CNet Svenska AB
peter.rosengren@cnet.se

Process Modelling and Monitoring Framework

The Process Modelling and Monitoring Framework is a web component based on well-known standards (B2MML, gbXML, OGC, x3d, BPMN). It enables modelling of all industrial aspects, such as definition of factory processes and factory resources (assets, equipment, sensors, buildings, actors, etc.) Through a web service API, data exchange between analytic tools and decision support systems is facilitated by the high level of simplicity, extensibility, interoperability and openness.

Contact: Marc Jentsch, Fraunhofer FIT
marc.jentsch@fit.fraunhofer.de

Secure Real-Time Event Brokering

To fully implement interoperability in real-time architectures, existing IoT Event and Messaging technologies (such as MQTT, AMQP, XMPP, STOMP) have been extended with real-time brokering and message translation services, allowing seamless integration of heterogeneous manufacturing components without costly re-configuration of processes.

Contact: Peter Rosengren, CNet Svenska AB
peter.rosengren@cnet.se

Simulation and Forecasting Toolkit

The web-based Simulation and Forecasting Toolkit (SFT) provides forecasts and predictions using live sensor data as well as historical data for both predictive maintenance and supply chain cases. The sensing layer integrates different sensors to support a Dynamic Reasoning Engine (DRE) and alarming services in production and logistics. The SFT feeds the Decision Support System and proposes actions. Hypothetical scenarios can also be tested.

Contact: Dimosthenis Ioannidis, CERTH ITI
djoannid@iti.gr

Visual Analytic Tool

The web-based Visual Analytic Tool (VA) provides advanced visualisations of live data from vibration sensors. The tool supports density based spatial clustering of applications with noise for detection of failures. The VA feeds the COMPOSITION Decision Support System with analytics results.

Contact: Dimosthenis Ioannidis, CERTH ITI
djoannid@iti.gr