Analysis of Digital Twins

Analysis of Digital Twins

The digital twin is a vision for a technology, originally conceived for NASA’s space programme, enabling industry to significantly improve the life-cycle management of physical assets. A digital twin is typically a system which collects data about a physical asset (such as a plant or a reservoir), continuously revises this data set through, e.g., updates reflecting changes to the asset’s structure and sensor data reflecting the physical asset’s state and uses this data to monitor and make predictions about the physical asset. The digital twin can be thought of as a three-layered structure: the data sources, an information layer and an insight layer. Industrial focus is today mainly on collecting data into shared, and increasingly structured, data sets which we think of as the information layer, and on providing dashboard-like insights into the system.

 

This project focuses on analysis support for digital twins, by building or combining tools which can leverage the information layer into insights. The purpose of these tools can be to reproduce and explain past events, to explore alternatives for decision making, to prepare for incidents or to optimize production. A central goal for this project is to combine semantics, behavioural and conceptual modelling techniques and analysis methods in the context of digital twins.  The methodological background for our work is a blend of ontology-based conceptual modelling techniques, formal methods and data-driven techniques for system analysis. In the last year this project has developed and implemented semantical programs to examine the relation of conceptual modelling techniques and formal methods, and integrated co-simulation units in a uniform framework.

Challenges

  • Understand the relation between digital twin infrastructure and asset models.
  • Understand the relation between time series and simulation results.
  • Develop experience with more complex digital twins.
  • Develop models for modular treatment of semantical programs.
  • Establish semantical programs as a research base for conceptual models in programming.

Our Approach

  • Develop a programming framework, SMOL, that seamlessly integrates domain knowledge with programs that can manipulate simulators
  • Implement methods for expressing “what-if” scenarios in SMOL.
  • Develop formal theory to extract needed digital twin simulator configurations from an asset model based on the information needs of users.
  • Develop prototype tool for programming with knowledge about traces/time series.
  • Collaborate with and disseminate results through publications and through, e.g., PeTwin

Results

Demo

Tutorial lectures (2x 90 minutes) at the PhD school of the International Conference on Theoretical Aspects of Computing (ICTAC 2022) in Tbilisi, Georigia, September 2022.

Documentation

https://smolang.org

Source Code

https://smolang.org

Publications

Eduard Kamburjan, Einar Broch Johnsen: Knowledge Structures Over Simulation Units. ANNSIM 2022: 78-89

Eduard Kamburjan, Sandro Rama Fiorini: On the Notion of Naturalness in Formal Modeling. The Logic of Software. A Tasting Menu of Formal Methods 2022: 264-289

Eduard Kamburjan, Vidar Norstein Klungre, Martin Giese: Never Mind the Semantic Gap: Modular, Lazy and Safe Loading of RDF Data. ESWC 2022: 200-216 [Best paper award]

Eduard Kamburjan, Vidar Norstein Klungre, Rudolf Schlatte, Silvia Lizeth Tapia Tarifa, David Cameron, Einar Broch Johnsen: Digital Twin Reconfiguration Using Asset Models. ISoLA (4) 2022: 71-88

Eduard Kamburjan, Crystal Chang Din, Rudolf Schlatte, Silvia Lizeth Tapia Tarifa, Einar Broch Johnsen: Twinning-by-Construction: Ensuring Correctness for Self-adaptive Digital Twins. ISoLA (1) 2022: 188-204

Rudolf Schlatte, Einar Broch Johnsen, Eduard Kamburjan, Silvia Lizeth Tapia Tarifa: Modeling and Analyzing Resource-Sensitive Actors: A Tutorial Introduction.COORDINATION 2021: 3-19

Eduard Kamburjan, Egor V. Kostylev: Type Checking Semantically Lifted Programs via Query Containment under Entailment Regimes. Description Logics 2021

Eduard Kamburjan, Vidar Norstein Klungre, Rudolf Schlatte, Einar Broch Johnsen, Martin Giese: Programming and Debugging with Semantically Lifted States. ESWC 2021: 126-142

Team

SIRIUS:

Eduard Kamburjan, Silvia Lizeth Tapia Tarifa, Rudolf Schlatte, Vidar Norstein Klungre, Martin Giese, Egor Kostylev, David Cameron, Einar Broch Johnsen,  a.o.

Partners

Acknowledgements

This work was partially supported by the SIRIUS Centre for Scalable Data Access (Research Council of Norway, project 237889) and partially by PeTwin project