Head of Reinforcement Learning for Humanoid Robotics

NEURA Robotics; Zürich; Vor Ort; CHF 140.000 - 240.000; Vollzeit

Zusammenfassung

NEURA Robotics in Zürich, Switzerland, seeks a Lead Reinforcement Learning Engineer to define and execute strategy for learning-based humanoid control. You will lead a team delivering locomotion, manipulation, and whole-body control for the 4NE1 platform.

You will translate cutting-edge research into robust, production-ready robotics software, collaborate with hardware, perception, and software teams, and drive sim-to-real transfer and deployment with strong safety and performance.

Qualifikationen

• Master's or Ph.D. in Robotics, Computer Science, ML, or AI.
• 8+ years of robotics, reinforcement learning, or machine learning experience with leadership.
• Proven track record deploying learning-based robotic systems in production environments.
• Strong ability to mentor multidisciplinary engineering teams.
• Hands-on experience with PyTorch, MuJoCo, ROS2, Isaac Sim, Python, C++.

Aufgaben

• Define and drive strategic roadmap for learning-based humanoid behaviors (RL, imitation, foundation-model robotics).
• Lead development and deployment of control policies for locomotion and whole-body motion on the 4NE1 platform.
• Establish scalable training, validation, and deployment frameworks and quality standards.
• Bridge research and production, turning cutting-edge methods into customer-facing capabilities.
• Drive sim-to-real transfer via domain randomization, system identification, and physics calibration.
• Mentor a high-performing RL engineering team and foster technical excellence.

Jobbeschreibung

NEURA Robotics in Zürich, Switzerland, seeks a Lead Reinforcement Learning Engineer to define and execute strategy for learning-based humanoid control. You will lead a team delivering locomotion, manipulation, and whole-body control for the 4NE1 platform.

You will translate cutting-edge research into robust, production-ready robotics software, collaborate with hardware, perception, and software teams, and drive sim-to-real transfer and deployment with strong safety and performance.