Description
The Formal Models Engine is responsible for creating the appropriate model representation of each pilot use case. These customized modular dynamic models, incorporate:
• intent-specific infrastructure subgraphs extracted from the KG representation of manufacturing assets and capabilities and representing the available manufacturing providers that satisfy the customer intents,
• necessary AI-based predictions of asset availability or demand fluctuation and asset or capability correlations, and
• Zero-X specifications like quantification and evaluation of circularity, and sustainability aspects. All the aforementioned parameters are taken under consideration by an automated pipeline employed to construct the corresponding formal model by determining its structure or formulation (e.g., DES, Flow Networks, MILP formulations, rule-based modeling) and tailoring it to the subsequent decision-making task or optimization algorithm leveraged on top of it.
Consequently, this component dynamically constructs models that enable the description of system behaviours, resource interactions, timing aspects, and control logic.
Main Innovations
- On-the-fly generation of models for supply chain optimization and matchmaking in the context of MaaS
- State space reduction through automatic generation of models based on data coming from the dataspaces and constraints from user intents
- Capturing temporal dynamics of supply chains
Licence
MIT
TRL
3 → 6
