Senior Robotic Controls & RL Simulation Engineer

Elysium Robotics; Greater London; Hybrid; GBP 70,000 - 90,000; Full time

Benefits offered by this job

Work from home flexibility

Support of US leadership

Opportunity to solve complex problems

Collaborate with elite talent

Job summary

Elysium Robotics is seeking a Senior Robotic Controls & Reinforcement Learning Simulation Engineer to join their innovative team in the UK. This role focuses on developing and optimizing control algorithms for advanced robotic systems using proprietary elastic actuators.

Your responsibilities will include training robust control policies, managing simulation infrastructures, and driving the performance of our bio-inspired robotic systems. Join us to tackle challenging deep-tech problems while enjoying a flexible hybrid work environment.

Qualifications

• Strong background in Reinforcement Learning and ML-based robotic control.
• Deep expertise in Variable Stiffness / Impedance Control.
• Experience with tendon driven robotic systems.

Responsibilities

• Develop robust control policies using Reinforcement Learning.
• Design control strategies for elastic muscle systems.
• Manage computational data infrastructure for simulations.
• Develop control for dexterous manipulation and rhythmic tasks.
• Develop scalable control strategies utilizing elastic actuators.
• Minimize sim-to-real gaps through constant assessment.

Skills

Reinforcement Learning

Variable Stiffness Control

System Identification

Tendon Driven Systems

Algorithm Complexity

Job description

Senior Robotic Controls & RL Simulation Engineer

Elysium Robotics is an early-stage company headquartered in Austin, Texas, developing revolutionary low-cost, highly capable robotic systems with dexterity and natural grace. Our proprietary muscle‑like elastic actuator technology addresses a decades‑old challenge in robotics, enabling automation for tasks current robots cannot do. These microfiber actuators enable real bio‑mimetic design opportunities based on the same muscle fiber actuation principles biology engineers with. We are building a world‑class team to revolutionize robotics.

Location: Initial hires will work from home with the expectation that a central office will be established in the coming months. This office will likely be in Central or East London, and a flexible hybrid schedule of in‑person and remote work will be established. There is also the possibility of locating in Cambridge, depending on various factors including team member locations.

Role Overview

As the Senior Robotic Controls & Reinforcement Learning Simulation Engineer you will develop the "brains" that drive early simulations of robotic systems built with our breakthrough elastic synthetic muscle actuators. You will focus on designing, training, and exploring control concepts using both emerging methods for variable stiffness control and advanced reinforcement learning within the VSim physics environment. Given the highly elastic nature of our synthetic muscles, your deep understanding of compliance, stiffness control, and antagonistic/reciprocal actuation will be critical to designing the reward functions and noise models that allow RL policies to successfully bridge the sim‑to‑real gap.

Our Collaborative Approach

You will work in a tight‑knit, two‑person simulation unit alongside our Senior Robotics Design & Simulation Engineer. While they architect the physical models, CAD, and environmental physics, you will lead the charge on how those systems are controlled, trained, and optimized. We expect both roles to creatively overlap and share the total scope of bringing our synthetic muscle technology to life in simulation. As our first two simulation hires, your combined skillsets will dictate the specific focus of our future team expansion.

Key Responsibilities

• RL & Policy Training: Use Reinforcement Learning and added noise models to train robust control policies that are tolerant of hardware uncertainty and the nuances of highly elastic components.
• Bio‑inspired Control Architectures: Design control strategies for reciprocal elastic muscle systems to independently manage torque and stiffness across complex robotic joints.
• Simulation Infrastructure: Develop and manage the computational data infrastructure required to launch hundreds of parallel simulation campaigns for rapid policy learning.
• Dynamics & Quasi‑statics: Develop control approaches for both quasi‑static tasks like dexterous manipulation and also explore algorithms for dynamically controlled resonance in rhythmic tasks (e.g., walking, swimming, flying, hammering, sawing, etc) within the simulation environment.
• High‑DOF Scalability: Develop control strategies that can eventually scale to systems utilizing hundreds or thousands of elastic actuators.
• Validation: Constantly assess the sim‑to‑real gap and work to minimize it.

Qualifications

• Strong background in Reinforcement Learning, ML‑based robotic control, and system identification.
• Deep expertise in Variable Stiffness / Impedance Control, active/passive compliance, and control of reciprocal / antagonistic actuation.
• Experience with tendon driven robotic systems.
• Ability to manage large‑scale computational workloads for parallel simulation campaigns.
• A keen eye for algorithmic complexity and computationally efficient methods.

Why Join Elysium Robotics UK?

• High Impact: Help build a new team in the UK from the ground up, supported by the leadership of our US HQ.
• Complex Challenges: Solve deep‑tech problems at the frontier of materials science, ML, robotics, and physics simulation.
• Flexibility: We offer a bias towards remote work while maintaining a key hub in the UK's premier tech corridors.
• Elite Talent: Work alongside senior, self‑directing engineers in a flat, fast‑moving organization.

We are an equal opportunity employer. If you require any reasonable adjustments during the application or interview process, please let us know.

We're funded by the UK's Advanced Research and Invention Agency.