Research Engineer, Reinforcement Learning

1X; Palo Alto (CA); On-site; USD 130,000 - 250,000; Full time

Job summary

A leading company in robotics is seeking a Research Engineer specializing in Reinforcement Learning. This role involves designing simulation environments for humanoid robots and addressing complex challenges in real-world applications. Successful candidates will have a strong background in Python programming and empirical research, contributing to innovative solutions in robotics.

Qualifications

• 4+ years of Python programming experience required.
• Strong empirical research abilities and performance bottlenecks spotting.
• Experience with tuning reward functions and deep RL.

Responsibilities

• Design simulation environments for robots and teach them new capabilities.
• Train robots for manipulation and locomotion tasks.
• Work with teams to ship RL policies to production.

Skills

Python programming

Empirical research abilities

Performance bottlenecks spotting

Tuning reward functions

Deep RL

Education

Advanced degree (MS or PhD) in Computer Science or related field

Tools

Mujoco

Pybullet

Isaac Sim

Job description

Join to apply for the Research Engineer, Reinforcement Learning role at 1X

1 month ago Be among the first 25 applicants

Join to apply for the Research Engineer, Reinforcement Learning role at 1X

Target start date: Immediately. Relocation provided.

Since its founding in 2015, 1X has been at the forefront of developing advanced humanoid robots designed for household use. Our mission is to create an abundant supply of labor via safe, intelligent humanoids. At 1X, you’ll own critical projects, tackle unsolved research problems, deliver great products to customers, and be rewarded based on merit and achievement.

We are looking for a Research Engineer in Simulation and Reinforcement Learning (RL). In this role, you will design simulation environments for NEO and teach NEO to learn new capabilities via RL algorithms. This enables our robots to be safe and robust in a variety of conditions.

Please only submit your application for one job posting at a time on the AI team. We will consider your application for all roles.

Responsibilities

1. Full-stack engineering, from data engineering to model architecture design to shipping polished products
2. Train NEO to do a diverse set of manipulation and locomotion tasks.
3. Close the sim2real gap between policies trained in simulation and real.
4. Work with controls, QA, and data collection teams to ship RL policies to the production fleet.
5. Deploy skills trained with RL into home environments
6. Getting general-purpose robots to work in the home is just about the hardest problem one can work on. We are looking for people with the courage to tackle unsolved technical challenges with an intense work ethic.
7. 4+ years of Python programming experience.
8. Strong empirical research abilities and a keen eye for spotting performance bottlenecks in RL training.
9. Experience with tuning reward functions, hyperparameters and exploration strategies to solve complex tasks with deep RL.
10. Experience with authoring environments and benchmarks in simulators like Mujoco, Pybullet, or Isaac Sim.

Nice to have

1. Advanced degree (MS or PhD) in Computer Science or related field
2. Published RL research in top ML conferences (NeurIPS, CoRL, RSS, ICML, etc.)
3. Have trained real-world quadruped or biped locomotion with RL
4. Robotics and control theory knowledge

Sample Projects

We encourage you to apply even if you do not meet every single qualification. If you have direct experience in solving one of the “sample projects” listed below, please let us know in your cover letter.

• Fuse separate RL policies (walking, running, standing up) into a single hybrid policy that smoothly transitions between modes.
• Speed up the simulator to enable faster training and evaluation.
• Design and implement infra for re-training models deployed in the real world using offline RL methods.
• Reduce the amount of “reward engineering” needed to solve long-horizon tasks by formulating general objectives like energy minimization, self-play, and data-driven reward functions.

Compensation

At 1X your work and results will be rewarded with a total rewards package consisting of a base salary, stock options and benefits. Base salary range is $130,000 to $250,000. Your actual salary will be based on your knowledge, skills and experience.

Location Policy

We believe the best work is done when collaborating and therefore require in-person presence in our office locations.