The M4ESTRO project, aimed at enhancing flexible, resilient, and reconfigurable value networks through trusted and transparent operations, has made significant progress across technical and scientific part during the first eighteen months of activities.
WP1: Resilient, Transparent, and Flexible Manufacturing Processes in Value Chains.
A deliverable D1.3 describing the Value chain Execution engine has been submitted. The Value Chain Execution Engine is a pivotal component of the M4ESTRO project, designed to enhance the efficiency and resilience of manufacturing value chains. This engine is capable of allocating internal capabilities and capacities obtained from IDS connectors developed in T1.3 and implementing the selected value-chain as outlined in T1.6.
The following major milestones were accomplished:
• MS01: a first initial set of indicators for internal and external disruptions in the supply chain has been defined. Firstly, the internal disruption indicators (T1.2) have been identified by looking at the possible disruption events along the logistic chain. Secondly, the external supply chain disruption indicators (T1.1) have been analysed and defined together with an initial data model, which will be utilised to serve both external and internal supply chain indicators in M4ESTRO
• MS02: it is reflecting advancements in prototypes for Resilience predictor tools: AI optimization Engine T1.4 Value Scoreboard T1.5 and Value chain engine T1.6.
Summarizing, the key tools developed include:
• IDS connector developed in T1.3 that will be used for peer-to-peer and trustworthy communication within the M4ESTRO ecosystem, allowing external stakeholders to exchange data in a secure and dynamic manner.
• A value execution engine in T1.6 to operationalize optimized chains and integrate dynamic inputs via IDS Connectors.
• A forecasting prototype to predict external disruptions in T1.1 using AI and rule-based analysis.
• A supply chain resilience prototype in T1.2 for real-time disruption assessment and reconfiguration suggestions
• A functioning prototype has been developed for the AI-Optimization Engine (T1.4) to explore intelligent supply chain planning using a multi-agent reinforcement learning (MARL) approach
• A fully functioning web-based user interface was developed under T1.5 to support interactions with the MARL-based supply chain decision system developed in T1.4.
WP2: Resilient Equipment, AI, and Trusted Data for Adaptive Manufacturing.
A major milestone, MS04 was accomplished with the implementation of a data infrastructure consisting of an IoT middleware SaaS. It acts as an integration hub for M4estro MaaS platform, connecting IT modules, manufacturing systems, equipment, and technologies into a unified IDS-compliant ecosystem.
Summarizing, the key tools developed include:
• A middleware-based data infrastructure established within T2.2 to facilitate real-time data transmission, featuring IDS-compliant interfaces and a dynamic ontology that embodies M4ESTRO semantics.
• A dynamic SLA system within T2.4 that was designed and finalized, shifting from a blockchain framework to a data space-based methodology to ensure compliance with the M4ESTRO architecture.
• A prototype for federated learning in distributed AI within T2.5 was completed, demonstrating privacy-preserving collaboration across supply chain participants.
WP3: Resilient Simulations to the Industrial Metaverse for Responsive Manufacturing
Summarizing, the key tools developed include:
• A preliminary version of the AAS representing an actor’s capabilities for a use case value chain has been created (T3.1).
• A sample application was implemented to showcase the practical realization of so-called Hybrid digital Twins and multi-actor simulations (T3.2 and T3.4).
• A demo that visualizes a 3D production cell for simulation purpose and what-if scenarios has been developed. (T3.3)
• A prototype was presented to illustrate the dynamic extension of the AAS through capabilities (T3.5).
WP4: Human-Centered Manufacturing Resilience and Sustainability.
Development of a prototype DSS has begun (T4.1). It is the result of the research on explainability techniques for the Decision Support System that identified post-hoc explainability and counterfactual generation, as the best approach. A framework was created to assess sustainability performance qualitatively using a maturity model and quantitatively by estimating MaaS CO2 emissions (T4.2).
WP5: Testing and Validation in Demo Cases.
Significant progress was made within WP5 to advance M4ESTRO project’s vision within the industrial context of each pilot.
Two chapters of the Trial Handbook have been developed and shared with the pilot coordinators and their technical partners from the three pilot projects to gather input on topics such as user needs, system requirements, and workflow processes (T5.1).
The consortium is engaged to ensure the alignment among partners and pilot use cases, improving the development of M4estro solutions since the architecture definition and components development.
