PC2: DTCOMPOSE

Overview

Digital Twins (DTs) have emerged as a means to interact with the physical world and explore what-if scenarios at different abstraction levels. However, most DTs are monolithic entities crafted in an ad hoc fashion for specific Physical Twins (PT).

To enable industrial scaling of this technology, we must make it possible to build DTs from modular parts, similar to how Gustave Eiffel revolutionized civil engineering in the late 19th century with the modular construction of the Eiffel Tower.

The challenges relate to how DTs can be modularized to allow composition at both design time and deployment time, at both syntactic (concrete form of interfaces) and semantic levels (ontology alignments). This modularization also concerns the services provided by DTs, including data processing and what-if exploration based on eg. machine learning.

Having a DT modeling a PT at a 1:1 scale is both a marketing claim and a scientific oxymoron. A model is always “wrong” with respect to reality, but some models can be helpful if you understand their distance from reality. This spans three dimensions: scale, fidelity, and uncertainty management. Scale and fidelity represent the level of abstraction (the scale of the map), while uncertainty management reflects confidence in DT attribute values (e.g., the river’s width is 10m ± 2m). Once identified within a DT, these can be propagated across DT boundaries at the federation level.

The project focuses on modular architecture, composition, interoperability, and managing scale, fidelity, and uncertainty with clear objectives: enabling modularization for flexible composition, orchestration at deployment, semantic interoperability through knowledge graphs, and proper uncertainty propagation across federated systems.

Associated Use Cases

Use cases in PC2 require interoperability as a critical feature, along with cases demonstrating modular approaches for building DTs or organizing DT federations.

Investigator & Project Partners

Principal Investigators:

Jean-Marc Jézéquel

Professor of Software Engineering at Université de Rennes/Inria

Jean-Marc Jézéquel is a Professor of Software Engineering at the University of Rennes and a member of the DiverSE team at IRISA/Inria, as well as a fellow of the Institut Universitaire de France (IUF). Since 2024 he is President of Informatics Europe. From 2012 to 2020 he was Director of IRISA, one of the largest public research labs in Informatics in France. In 2016 he received the Silver Medal from CNRS and in 2020 the IEEE/ACM MODELS career award. He was an invited professor at McGill University in 2022. His interests include model driven software engineering for digital twins with quality of service constraints, including security, reliability, performance, timeliness etc. He is the PI of the French-German MBDO project to help build Digital Twins for Industry 5.0 in a DevOps way. He is the author of 4 books and of more than 300 publications in international journals and conferences.

Jannik Laval

Associate Professor at Université Lumière Lyon 2

Jannik Laval is an Associate Professor in Software Science at Université Lumière Lyon 2 and a member of the DISP laboratory since 2015. He heads the laboratory’s research group names TOOLS, participates on the laboratory’s board, and is active within the national community (as part of the EDT working group at the Scilog research group, and organising conferences dedicated to digital twins). He obtained his HDR in 2020 at the University of Lyon, as well as a PhD in Computer Science in 2011 at the University of Lille. His research focuses primarily on the lifecycle management of software systems, systems of systems engineering, software engineering applied to cyber-physical systems, digital twins and their federation. As the author or co-author of over 55 journal articles and international conference papers, he develops research projects with national and international partners (including Polytechnique Montréal), both academic and industrial. He co-leads PC2 DTCOMPOSE with Jean-Marc Jézéquel.

Participating Partners:

Project Implementation

The project is structured into three complementary work packages, addressing modular architecture, composition, and semantic interoperability.