Aarhus University Seal


Nowadays, simulation is used for design space exploration, virtual testing or predictive maintenance for supporting early stage product decisions. Most importantly, real testing is ultimately used to assure product quality and certification. The aim of UPSIM is to enable companies to safely collaborate on simulations in a repeatable, reliable and robust manner and to implement simulations in a Credible Digital Twin setting as a strategic capability in order for them to become an important factor in quality, cost, time-to-market and overall competitiveness.

The UPSIM project funded by the Innovation Foundation Denmark in the ITEA3 framework to work towards reduction of physical tests by providing credible digital twins. In close collaboration with Agro Intelligence ApS, Aarhus University works towards credible digital twins able to incorporate sensory data at runtime in order to improve continuously.. The project has a full budget of 18M€ (~134MDKK) and is supported with 500k€ (~3.7MDKK) by the Danish Innovation Foundation. The project brings together expertise from 31 one partners from Austria, Germany, Denmark, the Netherlands, Romania and the United Kingdom.   

UPSIM is divided into 5 interrelated work packages to enable credible digital twins for production. The core problems addressed by UPSIM are:

  • Modelling & Simulation are of low economic value today as no measures are implemented for providing M&S credibility; this is directly associated to the limitation for virtual approval for reducing the amount of expensive real tests.
  • Modelling & Simulation responsibility is deeply-rooted in the domain-specific engineering departments, strictly leading to a bottom-up System Simulation approach and inherently conflicting with the verification and validation of cross-domain system functions; top-down approaches are mandatory.
  • Modelling & Simulation standards for the authoring and exchange of models, simulation and data are available and further developed, but the coordinated utilization along distributed product development, production and operation is mostly unclear and still missing.
  • Modelling & Simulation maturity is individual and hardly measurable.
  • Modelling & Simulation based on first principle models covers the intended system behaviour, but these models has to be extended for covering complex relations from real measurement data during engineering, production and operation.

      Aarhus University is specifically involved in the use case of developing credible digital twins for agricultural robots. In the process of moving towards autonomous robotic agricultural operations, several tasks rise to ensure continuous product reliability and performance without human interaction during operation. A common challenge when developing agricultural machinery is the ability to comply with the varying physical conditions across Europe, e.g. soil type and texture, topography, etc. Furthermore, the robots need to be compatible with a wide range of equipment and tools to be an integrated part of future farming. The agricultural robots consist of multiple sensors, software, electronics and mechanics. Tests is a crucial but time-consuming. By obtaining a validated digital twin of the system, the time to market can be reduced by performing tests in a virtual setting rather than performing physical tests that are often constrained by the availability of a test field, weather conditions and fully functional subsystems. By increasing the physics-based modelling, simulation and digital twin capabilities will provide valuable insights about the overall machine performance in the design phase as well as during field operation – the Figure below illustrates the realization and work panned by the involved partners    

Key Elements of UPSIM

  1. Following a strict model-based development scheme, system requirements are linked to functions and architectural system elements, ensuring traceability and evidence in system and digital twin design. AU will work on system simulation architectures allowing to dynamically distribute modelling and simulation efforts to maximise the performance and efficiency.
  2. Collect and manage real-world measurements and use them in the Digital Twin setting. This will enable us to reuse data from previous runs and utilise the information to make validation of vehicle models and environmental modes during runtime.
  3. Work towards dynamic integration of new models in the Digital Twin setting. This includes integration of partial models during runtime allowing rapidly improved and more complete models to be used as soon as they become available. Furthermore, we will develop approaches that allow ad-hoc deployment of established models and techniques towards the physical testbed were available. This will enhance the agility of development processes new products based on the experiences gained from the Digital Twin. The approaches developed will be generic but demonstrated on the off-road vehicles in agricultural settings.