From Dortmund to Singapore: research for secure AI systems worldwide

As part of the Walter Benjamin Program of the German Research Foundation (DFG), Dr. Junjie Shi, a postdoctoral researcher from Prof. Jian-Jia Chen's working group at the Department of Computer Science, is working on the safe integration of machine learning in real-time systems. The research project is entitled "Balancing Uncertainty and Determinism: Safely Integrating Machine Learning in Real-Time Systems".

The aim of the project is to develop new methods to enable machine learning (ML) to be used reliably in safety-critical and resource-limited embedded systems - for example in autonomous vehicles, industrial control systems or medical devices. The focus here is on three aspects: predictable ML execution over time, federated learning via distributed platforms and the use of ML techniques to improve planning and verification in real-time systems.

For his research, the young scientist is collaborating with Prof. Tulika Mitra's group at the National University of Singapore (NUS). With its internationally leading expertise in the field of embedded systems and hardware-software co-design, the NUS offers an ideal environment for the project.

Junjie Shi's stay in Singapore is not only intended to strengthen his scientific independence, but also to promote exchange between Europe and Asia. In the long term, this should lead to sustainable cooperation between TU Dortmund University, NUS and other international partners.