TwinTraSys

Holistic and dynamic mapping of transport system behaviour

In-plant transport systems must be flexible, as material flows can change at short notice. Various adjustments can be made to adapt the transport system to the requirements. For example, the use of resources for individual tasks can be varied. The planning and control of these systems is therefore characterised by a high degree of complexity and dynamism. The holistic system behaviour can be mapped through the use of digital twins. For this reason, the project with the research of model- and learning-based methods for the conception and prototypical implementation of digital twins for internal transport systems.

Conception of the digital twin

In the first step, the system architecture, the individual functions and the interfaces of the software services to other systems are developed and harmonised. The findings from process analyses of the project partners' use cases serve as the basis. For this purpose, a methodology for analysing logistics processes with a focus on creating digital twins was developed within the project.

The results of the process analysis are then translated into a technical concept for all services. This means that the individual functions and interfaces of the services are worked out and harmonised. For example, algorithms for data forecasting are developed in detail. A concept for analysing possible courses of action is also being developed. In addition to the effective identification of sensible alternative courses of action, the preparation of the results and the transparent presentation of the decision-making process play an important role. The results from simulation runs serve as a basis. For this purpose, generic simulation modules are designed that allow automatic model generation and parameterisation.

Implementation and evaluation

In addition to the software implementation of services for the simulation and parameterisation of simulation models within a simulation framework, algorithms for data forecasting and decision-making are also implemented. To test the functionality, the interactions between the individual services are tested in special test environments.

Finally, the digital twin will be functionally tested in a real environment. In addition to evaluating it and demonstrating its usability, the potential of the digital twin will also be analysed, taking into account economic and process-related aspects. To this end, the digital twin will be implemented at the project partners and in the TZ PULS model factory (Landshut University of Applied Sciences). Ongoing monitoring of the reference systems and iterative optimisation measures will ultimately make it possible to guarantee its functionality and usability.