Bart Rosier – speaker at Advanced Humanoid Forum 2027 Munich

Bart Rosier

Vesoma · Industrial Humanoid Platform

Vesoma

Industrial HumanoidsReinforcement LearningPhysical AI

About this Speaker

Bart Rosier represents Vesoma, a Munich-based company building an industrial humanoid platform. Vesoma emerged from stealth in September 2026 with a team of 60+ people across Munich and Limassol (Cyprus), and a distinctively different technical bet: instead of training humanoids primarily from human demonstrations, Vesoma builds a self-learning physical agent based on reinforcement learning.

The first commercial platform, Vesoma 1, is planned for unveiling in early 2027. Its stated target is not a research demo, but a robot that can run a full working shift without engineer intervention — a claim that, if delivered, would move the industry past one of its central credibility problems. Target industries are manufacturing, logistics and warehousing. The Chief AI Officer is Martin Riedmiller, previously of Google DeepMind.

Why this perspective matters

  • Alternative Technical Bet: Vesoma is one of the few European humanoid companies building on reinforcement learning and body-behavior co-evolution. That makes their perspective genuinely different from the demonstration-heavy mainstream.
  • Full-Shift Reliability as the Bar: Vesoma sets the benchmark at running a whole shift without an engineer — a much stricter target than "works in a demo," and one that speaks directly to what industrial buyers care about.
  • Munich-Based, European Supply: A 60+ person team in Munich with top-tier AI leadership (ex-DeepMind) is directly relevant to Europe's ability to build sovereign humanoid capability.
  • Industrial Focus: Vesoma targets manufacturing, logistics and warehousing from day one — not consumer or service robotics. The forum's audience is exactly the customer segment Vesoma is designing for.

Talk at Advanced Humanoid Forum 2027

Topic: "Self-Learning Humanoids – Why Reinforcement Learning Changes the Industrial Deployment Question"

January 20–21, 2027 · exact time to be announced

The session examines Vesoma's technical route in detail: why the team chose reinforcement learning plus co-evolution of body and behavior, how this differs from the demonstration-driven approach used by most competitors, and what it means for reliability, cost per task, and time-to-productivity in industrial settings. The talk is aimed at industrial decision-makers who need to separate short-lived demos from systems that can actually be operated.

Areas of Expertise

Industrial HumanoidsReinforcement LearningPhysical AISelf-Learning SystemsManufacturing & LogisticsHumanoid ReliabilityEuropean Robotics

More Speakers at Advanced Humanoid Forum 2027

Attend Advanced Humanoid Forum 2027

January 20–21, 2027 · Midnightbazar Munich · Two stages, Workshop Area and the live Application Area powered by Schnaithmann