Join the team

Reinforcement-Learning Simulation

Build the simulation environments, synthetic-data factory, and Real2Sim strategies that close the sim-to-real gap.

Responsibilities

Build and maintain robotic-manipulation simulation environments using Isaac Lab or Genesis, while managing the accuracy of physical parameters.
Design a large-scale automated synthetic-data production pipeline supporting diverse randomisation of scenes, objects, and lighting.
Research domain-randomisation and Real2Sim strategies, quantify the system gap between simulation and the real world, and reduce that gap.
Build reinforcement-learning environments supporting thousands of parallel instances, and design reward functions and curriculum-learning schemes.
Establish a 3D asset pipeline supporting USD, URDF, and MJCF, covering the complete process from object scanning to annotation of physical properties.
Collaborate with the pre-training, post-training, and hardware teams to connect simulation data with model training in an end-to-end pipeline.

Qualifications

Master’s or Ph.D. degree in a relevant field, with familiarity with Isaac Lab, MuJoCo, or comparable platforms.
Hands-on project experience with large-scale parallel simulation training or synthetic-data production.
Understanding of robot kinematics and dynamics, with familiarity with URDF and MJCF formats.
Practical sim-to-real experience is strongly preferred.

Location: Hangzhou, Beijing, Switzerland. Send your CV to info@awomo.ch. Suggested subject line: “Name + Position” — the apply button fills in the role for you, so just replace “Name”.