Humanoid Locomotion Reinforcement Learning Engineer

Generative Bionics; Uri; In loco; EUR 70.000 - 120.000; Tempo pieno

Descrizione del lavoro

Generative Bionics cerca un/una Humanoid Locomotion & Reinforcement Learning Engineer per sviluppare locomozione e movimento del corpo completo per la sua piattaforma robotica umanoide. Lavorerai all’intersezione tra robotica, ML e controllo, progettando soluzioni RL e guidando l’intero ciclo d’implementazione, dalla simulazione al training delle policy e al deployment su robot reali.

Collaborerai strettamente con i team Meccanico, Elettronica, Percezione, Controlli e IA per integrare le

Competenze

• Titoli avanzati in robotica, controllo o ML.
• Esperienza in RL per robotica umanoide o legged.
• Conoscenze di simulazione robotica e controllo.

Mansioni

• Sviluppare policy RL per locomozione e controllo equilibrio.
• Progettare pipeline di generazione movimenti e retargeting.
• Costruire modelli robot/attuatori in simulazione.
• Implementare transfer sim-to-real e deployment su robot reali.
• Collaborare con team meccanico, elettronico, percezione e IA.

Conoscenze

Analytical thinking

Team collaboration

Multidisciplinary teamwork

Formazione

Master’s or PhD in Robotics/Control/ML/CS

Strumenti

Python

C++

PyTorch

MuJoCo

Isaac Sim/Lab

Descrizione del lavoro

Descrizione dell’offerta di lavoro

Location: Via Melen 83, 16152 Genoa, Italy

Contract: Full-time, Permanent

About Us

Generative Bionics is a deep-tech company building humanoid robot platforms to deploy human-centered Physical AI. We design intelligent, capable machines that work alongside people in real-world environments — developed in Genova, Italy.

Role

We are looking for a talented and driven Humanoid Locomotion & Reinforcement Learning Engineer to develop advanced locomotion and whole-body motion capabilities for our humanoid robot platform. In this role, you will work at the intersection of robotics, machine learning, and control systems, designing and deploying reinforcement learning-based solutions that enable robust, dynamic, and adaptive robot behavior. You will contribute to the full development pipeline, from simulation and policy training to sim-to-real transfer and deployment on physical robots.

Responsibilities

• Develop and train reinforcement learning policies for humanoid locomotion, balance control, and whole-body motion;


• Design motion generation, imitation learning, and motion retargeting pipelines using demonstrations, motion capture data, and reference trajectories;


• Build and maintain accurate robot, actuator, and contact models using simulation environments such as Isaac Lab, Isaac Sim, MuJoCo, or equivalent platforms;


• Develop domain randomization, system identification, and adaptation techniques to improve sim-to-real transfer performance;


• Deploy, validate, and optimize learned control policies on physical robots using Python and C++;


• Implement monitoring, fall detection, recovery strategies, and policy validation mechanisms to ensure safe robot operation;


• Analyze performance through simulation results, telemetry, robot logs, and experimental testing;


• Collaborate closely with Mechanical, Electronics, Perception, Controls, and AI teams to integrate locomotion capabilities into the humanoid platform;

Requirements

• Master’s degree or PhD in Robotics, Control Engineering, Machine Learning, Computer Science, or a related field;


• Experience developing and applying reinforcement learning techniques to humanoid, legged, or whole-body robotic systems;


• Strong knowledge of robot kinematics, dynamics, contact modeling, state estimation, and feedback control systems;


• Experience working with robotics simulation platforms such as Isaac Lab, Isaac Sim, MuJoCo, or equivalent environments;


• Knowledge of deep reinforcement learning, imitation learning, motion priors, or learning-based control approaches;


• Strong Python programming skills and practical experience with C++ for real-time robotic applications;


• Experience with PyTorch or equivalent machine learning frameworks;


• Experience developing, testing, and debugging software on physical robotic systems;


• Familiarity with Linux, Git, and software development best practices;


• Strong analytical and problem-solving skills, with the ability to work effectively in multidisciplinary teams;

Valued Extras

• Experience generating, retargeting, blending, and adapting motion priors from motion capture datasets, demonstrations, animation assets, or learned motion datasets;


• Knowledge of whole-body control, model predictive control (MPC), trajectory optimization, inverse dynamics, or hierarchical control architectures;


• Experience with sim-to-real methodologies, loco-manipulation, or contact-rich robotic behaviors;


• Familiarity with fall prevention, disturbance rejection, recovery strategies, and safe policy execution techniques;


• Publications in robotics, machine learning, or control systems conferences and journals;


• Contributions to open-source robotics projects or demonstrated personal robotics projects;

We Offer

• The opportunity to contribute to the development of cutting-edge humanoid robotic systems;


• Work on challenging robotics and Physical AI problems with direct real-world impact;


• A stimulating and informal work environment alongside highly skilled technical and research teams;


• Employment contract under the Italian Metalworking Collective Labor Agreement (CCNL Metalmeccanico), commensurate with experience;


• Concrete opportunities for professional growth;

Disclaimer

We are proud to be an Equal Opportunity Employer. We evaluate all qualified applicants solely on the basis of merit and business needs, without distinction or discrimination based on gender, race, color, ethnic or social origin, age, religion, sexual orientation, gender identity, disability, or any other characteristic protected by law.