Research

Overview

My research focuses on the validation of automotive radar perception systems, with an emphasis on robustness, reliability, and safety-critical behavior in advanced driver assistance and automated driving.

A central challenge in perception system development is the rarity of critical edge cases—scenarios that occur infrequently in real-world data but dominate safety risk. My work addresses this challenge by combining synthetic data generation, system-level validation, and engineering-driven evaluation methodologies.

I aim to bridge academic research and industrial practice, ensuring that research contributions translate into measurable improvements in real-world systems.

PhD research

Doctorate in Electrical Engineering (ongoing)
RWTH Aachen University

Topic

Validation of radar perception systems through synthetic edge-case generation

Research questions

• How can synthetic data be used to systematically expose rare but safety-critical perception failures?
• Which validation metrics meaningfully reflect real-world perception robustness?
• How can radar-specific characteristics be preserved and controlled in synthetic scenarios?
• How can validation methods scale to industrial development processes?

Research directions

Synthetic data & edge cases

• Generation of radar-specific synthetic scenes and point clouds
• Controlled variation of scene parameters to uncover failure modes
• Coverage-driven validation strategies beyond mileage-based testing

Radar & multi-sensor perception

• Radar signal processing and object perception
• Tracking and classification in complex traffic scenarios
• Interaction between radar and complementary sensing modalities

System validation & robustness

• Degradation detection and performance monitoring
• Robustness assessment under environmental and sensor uncertainties
• Validation methodologies aligned with safety and regulatory needs

Explainable & dependable systems

• Interpretability of perception outputs and failure modes
• Traceability between system requirements, data, and behavior
• Engineering approaches for dependable AI-based perception

Publications & intellectual property

• Generative Adversarial Synthesis of Radar Point Cloud Scenes
  IEEE MTT / ICMIM 2024

• Author and co-author of multiple granted patents in automotive radar perception, including sensor synchronization, degradation detection, and system operation methods.

A complete list of publications and patents is available in the downloadable CV.

Research environment & collaboration

My research is conducted in close collaboration with industry partners, allowing for continuous feedback between theoretical methods and practical constraints.

I am particularly interested in collaborations related to:

• perception system validation and benchmarking
• synthetic data generation for safety assessment
• industry–academia knowledge transfer

For prospective collaborators

If you are interested in collaborating, discussing research ideas, or exploring joint projects, please reach out via the Contact page.

I am especially open to discussions that connect theoretical rigor with practical impact.

This page reflects my current research focus and will evolve as the work progresses.