From tinkering with LEGO robots to PhD research at Stanford. Here's my story.
2012–2019
FIRST LEGO League
International robotics competition
7 years; 2× 1st Place Programming; highest robot-game score of any German team (2019); World Festival, St. Louis (2017).
My fascination with robotics started right here, programming LEGO Mindstorms robots to move autonomously across the game field. ▶ watch
2020–2023
BSc, Engineering Cybernetics
University of Stuttgart · 1.2 / 1.0 (Honours)
Control theory & dynamical systems. Thesis: automatic landing of a model aircraft via fuzzy control (1.0).
Wanting to keep heading toward intelligent systems, I picked Engineering Cybernetics for its control-theory-heavy core, the groundwork I wanted for robotics.
2023–2024
Engineering Intern, R&D
Mercedes-Benz AG
Slip-control algorithm for clutch-based torque distribution in an electric drivetrain.
Joining the R&D drivetrain-control team at Mercedes-Benz was my first real test of developing control algorithms in an industry setting.
2024
Student Software Engineer
Vector Informatik GmbH
Compiler validation with fuzz-testing (.NET / C#).
At Vector I focused on the software side for a while, getting a solid stretch of real-world engineering.
2024–2026
MSc, Robotics, Systems & Control
ETH Zürich · 5.83 / 6.0 (with distinction)
Research at ETH's Robotic Systems Lab and UZH's Robotics & Perception Group. Master's thesis (6.0): Bridging Agentic Reasoning and Structured Execution: A Behavior-Tree VLM Framework and a Fidelity-Ladder Benchmark for Embodied Agents.
At ETH I finally came back fully to robotics, getting into learning-based methods and bridging classical control with machine learning.
2024–2025
BMW Fastlane Scholar
BMW AG
CV street-model architectures for autonomous driving; RL / VLA finetuning for pick-and-place on a production use-case.
BMW's Fastlane scholarship gave me a close look at autonomous driving and manipulation research in industry.
2025–2026
Visiting Researcher
Stanford University · Autonomous Systems Lab
Agentic orchestration and full-stack embodied question answering for mobile robots on the DARPA TIAMAT project.
Master's thesis (6.0): Bridging Agentic Reasoning and Structured Execution: A Behavior-Tree VLM Framework and a Fidelity-Ladder Benchmark for Embodied Agents.
I had always hoped to do my master's thesis in the US, and this is where it happened.
2026–present
PhD, Robotics & Autonomy
Stanford University · ASL (Marco Pavone)
In collaboration with UC Berkeley (S. Shankar Sastry's group). Task-level autonomy & world models; DARPA TIAMAT.
Building on the momentum from my time here, staying on for the PhD was an easy call: a top environment to go all in on embodied-intelligence research, with people I enjoy working with.
// publications
Under review · CoRL 2026
FARM: Find Anything using Relational Spatial Memory
Siming He, Leo Huang, Adam Lilja, Fabio Hübel, Jonas Frey, Marco Pavone, S. Shankar Sastry, Jitendra Malik, Claire Tomlin