Hands-On Robot Manipulation Researcher (Real Hardware)

Menlo; Singapore; On-site; SGD 120,000 - 180,000; Full time

Job summary

Menlo is seeking a Researcher, Manipulation to advance the Asimov humanoid platform. You will drive progress in grasping, dexterous and contact-rich tasks, coordinating arms, torso, and legs to solve challenges in real hardware.

You will design, train, and deploy RL policies, improve sim2real transfer, and collaborate with locomotion, controls, and hardware teams while contributing to open-source efforts.

Qualifications

• Hands-on experience with reinforcement learning for manipulation and dexterous control.
• Track record of deploying manipulation policies on real robots, not only in simulation.
• Strong programming skills in Python and ROS2-based control stacks.
• Proficiency with physics simulators (MuJoCo, Isaac) including contact modeling and domain randomization.

Responsibilities

• Design, train, and ship RL policies for manipulation, grasping, and dexterous tasks on Asimov.
• Extend RL work toward whole-body manipulation with balance and reaching integrated.
• Own the sim2real pipeline for manipulation to ensure transfer to hardware.
• Design rewards and training environments, including randomization strategies in MuJoCo.

Skills

Reinforcement Learning

Manipulation

Python

MuJoCo

ROS2

Tools

MuJoCo

ROS2

Isaac

Job description

Menlo is seeking a Researcher, Manipulation to advance the Asimov humanoid platform. You will drive progress in grasping, dexterous and contact-rich tasks, coordinating arms, torso, and legs to solve challenges in real hardware.

You will design, train, and deploy RL policies, improve sim2real transfer, and collaborate with locomotion, controls, and hardware teams while contributing to open-source efforts.