Tim-Louis Lewejohann, M. Sc.

Research interests

My research focuses on improving the reliability and fault tolerance of hardware accelerators for neural networks, particularly in resource-constrained edge systems. I investigate how hardware faults affect neural network inference and develop methods to identify and protect vulnerable components. The goal is to enable efficient and dependable AI systems for safety-critical applications, such as autonomous driving and space exploration.

• Reliable and Fault-Tolerant AI Accelerators

• Fault Injection and Vulnerability Analysis

• Cross-Layer Resilience and Selective Hardware Hardening

Career

since 2024:

Research Assistant at the Institute of Computer Engineering, University of Lübeck

2022-2024:

MSc.   Robotics and Autonomous Systems, Universität zu Lübeck

2018-2022:

BSc.   Robotics and Autonomous Systems, Universität zu Lübeck

Theses

I supervise Bachelor's and Master's theses in the areas of reliable AI, neural network accelerators, and fault-tolerant computing. Possible topics include hardware fault injection, vulnerability analysis of neural networks, FPGA-based accelerator design, and techniques for improving the resilience of AI systems.

Students interested in these areas are welcome to contact me via email, ideally with a brief description of their interests, relevant experience, and whether they are looking for a Bachelor's or Master's thesis. Individual topics can be discussed and adapted to the student's background and interests.