Auto-Disco

Auto-Disco: Automated Discovery of Model Deficiency and Model Update in Process Industrial Digital Twins

The proposed research aims to generate and deepen the knowledge and to produce methods and tools for digital twins of process industrial systems that enable the autonomy and artificial intelligence to keep a digital twin updated during the real life cycle of the system. The competence at LTU will thus be further strengthened in the area of ​​model learning, control and monitoring for process industrial systems using digital twins.

The control technology group at LTU has a long history of process industry-related research with a focus on resource efficiency and reliability in industrial processes. Since 2005, the Automatic Control group has been involved in shaping the research within ProcessIT Innovations as an important network.

In recent years, energy-related applications from the data center area and the heating and cooling sector have been added, considering their similarities with the process industry sector and contributing to the DHC+ Technology Platform and the European Technology and Innovation Platform for Renewable Heating and Cooling as experts.

The focus of the research is therefore on research methods that can automatically detect model deficiencies and autonomously update the models. In this way, engineers get tools to create and use digital twins more effectively. With the funding, we will recruit a postdoctoral fellow for a period of two years to carry out this research.

Funded by: Kempe Scholarships

Contact: Professor Wolfgang Birk