Generic Digital Twins
Semantic modelling of heterogeneous entities and processes, a modular core for data ingestion and synchronisation, and common interfaces across application domains.
Interoperability · Abstraction · Data modelsTowards the next generation of intelligent Digital Twins.
A research programme investigating how modular, domain-independent Digital Twins can combine Generative AI and Multi-Agent Systems to model complex environments, explore new scenarios and support adaptive decision-making.
Digital Twin platforms often remain tied to one application domain, using heterogeneous data models and interfaces that are difficult to reuse. SCULPT explores a more general approach: a layered architecture with shared semantic representations and interoperable components.
The funded 2026–2028 start-up phase focuses on architectural design, core modules, initial integration of generative models and agents, and a proof-of-concept evaluation. The longer-term research ambition is broader, but is not presented here as already funded.
The technical research combines semantic Digital Twin engineering, scenario generation and coordinated decision-making.
Semantic modelling of heterogeneous entities and processes, a modular core for data ingestion and synchronisation, and common interfaces across application domains.
Interoperability · Abstraction · Data modelsExploration of generative models to produce plausible what-if scenarios, support dynamic simulations and help Digital Twins respond to changing environments.
LLMs · Scenario generation · SimulationResearch into specialist agents, interaction protocols and orchestration strategies supporting collaborative reasoning and adaptive system behaviour.
Agent roles · Coordination · Decision supportDefine use cases, requirements and the modular system blueprint.
Develop the first Generic Digital Twin core and experimental GenAI–agent integration.
Evaluate feasibility through a restricted pilot and initial performance indicators.
Disseminate early findings and prepare subsequent competitive research proposals.
These activities describe planned project objectives, not results already achieved.
SCULPT brings together Computer Science, Software Engineering and Intelligent Systems expertise, building on M.O.D.A.L.’s research in Digital Twins, AI and data-driven modelling.
Meet the research group →Explore SCULPT alongside the laboratory's related work in Generative AI and Digital Twin research.