I am Victor Lutz, a Robotics & ML Engineer currenly based in France. I am interested in Legged Robotics, Reinforcement Learning, Simulation, Manipulation Policies (Flow/Diffusion, VLA), and Model Predictive Control.
From November 2024 to November 2025, I was working as a Research Engineer in Humanoid Robotics at LAAS-CNRS Gepetto in Toulouse, France. My work focused on the control software of legged robots with parallel mechanisms, with both Reinforcement Learning and Optimal Control.
Prior to that (2024), I was interning at LAAS-CNRS Gepetto, working on the same topics. Even prior to that (2023), I was interning in PAL Robotics, working on robot simulation.
My current goal is to apply my skills to allow robots doing useful work in the real world, using cutting-edge techniques. As of March 2026, I am looking for new opportunities.
Selected Publications
Control of Humanoid Robots with Parallel Mechanisms using Differential Actuation Models Victor Lutz, Ludovic De Matteïs, Virgile Batto, Nicolas Mansard IEEE ICRA 2026, Vienna, Austria. arXiv:2503.22459 · Google Scholar
Selected Projects
Humanoid Control with Parallel Mechanisms — LAAS-CNRS Gepetto, 2024–2025
A complete framework for controlling humanoid robots with closed-kinematic actuators (parallel knee and ankle mechanisms). Introduces a compact differential analytical model enabling efficient trajectory optimization and RL deployment with impedance transfer directly on hardware.
Paper (arXiv:2503.22459) · More projects
News
Mar. 2026 — Graduated from the Deep Reinforcement Learning Nanodegree (Udacity / NVIDIA Institute) !
Jan. 2026 — Paper accepted at IEEE ICRA 2026: Control of Humanoid Robots with Parallel Mechanisms using Differential Actuation Models !