The development of reliable and agile digital twins of high-tech systems and materials is key to enable shorter time-to-market, zero-defect and flexible manufacturing systems with accurate predictive maintenance. This crucial development is currently hampered by the lack of synergy between model-based engineering and data-driven/artificial intelligence approaches. The DIGITAL TWIN program will develop key-enabling technologies for full digitization of the value chain of high-tech systems and materials by the integration of data-driven learning approaches and model-based engineering methods.
The NWO AES Perspectief DIGITAL TWIN Program is a five-year comprehensive research program on the development of digital twins and digital twinning methods, financed by the Dutch Research Council (NWO) within the domain of Applied and Engineering Sciences (AES). This collaborative programme involves six universities: University of Groningen, Eindhoven University of Technology, TU Delft, University of Twente, Leiden University and Tilburg University and twelve industrial partners.
Within DIGITAL TWIN, project 3 ('P3') is led by Eindhoven University of Technology (TU/e) and Tilburg University (TiU), and addresses
methodology and tool support for effective digital twinning. Creating dedicated digital twin models involves the combination of heterogeneous models from different engineering disciplines - used to model different DIGITAL TWIN components - and modeling of the interactions between these different component models. The challenges addressed in this project are (1) ensuring consistency between the involved heterogeneous models, as well as early identification of inconsistencies in order to ensure a trustworthy digital twin; and (2) the efficient creation of digital twins using model-driven software engineering (MDSE) techniques.
Three independent but related work packages comprise DIGITAL TWIN's project P3. Two of these comprise the two PhD positions at TU/e:
-
PhD position on digital twin model consistency (based at TU/e). In this project, you will address the fundamental research question of how to discover and manage consistency relations between models across engineering domains. Models developed by engineers from various disciplines need to be aligned and made consistent, and the engineers need to be informed of any inconsistencies. For this purpose, an existing model management framework needs to be extended with consistency checking. The position will require the development of a new methodology for discovering, constructing, and checking cross-domain consistency rules. It will also require the development of good tool support that covers interoperability with existing languages and tools in the industry, allowing engineers to use existing tooling and not be needlessly bothered by consistency tracking functionality.
-
PhD position on digital twin orchestration (joint PhD between TU/e and TiU, but based at TU/e). In this project, you will address the fundamental research question of how to define languages that allow the description of the orchestration between the heterogeneous models in order to create and adapt digital twins. MDSE techniques can be used with existing heterogeneous models to define the orchestration of their interaction, i.e., to integrate them into an executable artefact. This will require a framework that supports (1) the integration of new and existing heterogeneous models via code encapsulating the models, (2) exchange of data between the models, and (3) feeding them with real-life data from the physical artefact that the digital twin is coupled to. The goal will be to reuse an existing framework (to be investigated), and extend it efficiently based on the use of MDSE techniques. An important part of the project will be the investigation of suitable language elements (concepts) that need to be part of a domain specific language for specifying the creation of digital twins. The resulting Domain Specific Language with model transformations and code generators needs to support the above mentioned requirements (1) through (3), mapping language elements to the APIs of the selected framework.
Project EmbeddingApart from these two PhD positions based at TU/e, a third, separate PhD position on the related topic of
digital twin dynamic consistency will be advertised separately by TiU and based at TiU.
For all three projects, industrial digital twin cases as provided by the industrial project partners (Philips Personal Care and StamiCarbon) will be addressed at the 'proof of concept' level.
The sub-projects, hence, the PhD positions are independent but the PhD researchers are expected to cooperate on the common project goals. As a result, regular visits to or working from TiU 1-2 days/week and research visits to the industrial partners for case studies may be required.