Senior Robotics Simulation Engineer

Automated Architecture (AUAR); Greater London; On-site; GBP 70,000 - 110,000; Full time

Benefits offered by this job

Health and dental insurance via Bupa

11,000 sq ft R&D facility in London (E

Share options

Job summary

AUAR, a London-based physical AI and robotics company automating homebuilding, seeks a Senior Robotics Simulation Engineer to own and advance the simulation capability for our MicroFactory. You will lead the physics-based environment in NVIDIA Isaac Sim, maintain digital twins, CAD/USD asset pipelines and ROS 2 integration.

You’ll define architecture, ensure fidelity balance, collaborate with robotics software, controls and mechanical engineering, and build scalable tooling that the team can

Qualifications

• You've built robotics simulation environments in NVIDIA Isaac Sim.
• Strong background in physics-based simulation and ROS 2 integration.
• Excellent Python, Linux, Docker and Git skills.
• Ability to design scalable simulation infrastructure for teams.
• Good judgment on fidelity and trade-offs between models and real systems.
• Cross-discipline collaboration with controls, mechanical, and software teams.

Responsibilities

• Own and evolve Isaac Sim environment for robotics development.
• Maintain digital twins, asset pipelines and USD workflows.
• Tune physics like contacts, friction, joints and actuators.
• Develop ROS 2 interfaces for simulated and real systems.
• Create virtual development scenarios and automated tests in CI.
• Identify sim-to-real gaps and drive fixes.
• Generate synthetic data for CV training with domain randomisation.

Skills

NVIDIA Isaac Sim

Python

Linux

Docker

Git

ROS 2

CAD/3D assets

Architecture

Tools

Isaac Sim environment

Job description

We're automating homebuilding. Construction is one of the least automated industries in the world, and homes are still built largely by hand, in slow, expensive and fragmented processes. AUAR is a physical AI and robotics company changing that.

Our software platform, MasterBuilder, turns a building design into the instructions needed to manufacture a home's timber structure. Our robotic MicroFactory then deploys directly onto a homebuilder's site and produces the wall, floor and roof panels they need, on demand. It arrives in a shipping container, sits on a footprint of max 30 by 30 ft and runs off generators on a gravel pad: no grid connection, no permanent site works. We own and run the equipment, our customers buy the panels it produces, not the machine.

We're past pure R&D and shipping our first commercial MicroFactories this year, across the UK, Europe and North America. Backed by Planet A, PT1, Shadow Ventures and ABB Robotics, we're a team of around 35 engineers, roboticists, operators and designers working from an 11,000 sq ft R&D and demonstration facility in London. The systems we build now will decide how homes are manufactured at scale.

The role

We're looking for a Senior Robotics Simulation Engineer to own and advance the simulation capability for our robotic MicroFactory.

Simulation is how we develop an increasingly complex robotic manufacturing system without being blocked on one physical machine. You'll own our physics-based environment in NVIDIA Isaac Sim and everything around it: digital twins of the robotic cell, CAD and USD asset pipelines, physics and control modelling, ROS 2 integration and automated test scenarios. The hardware keeps changing underneath you, so keeping the digital and physical systems aligned is permanent, not a setup phase.

This is a senior IC role with real technical ownership. You'll define the engineering approach, make the calls on fidelity and architecture, and work with robotics software, controls, integration, test and mechanical engineering to find where simulation actually has leverage. You'll also make it something the team can use, not something only you can run.

Problems you'll work on

• Simulation architecture - Own and evolve our Isaac Sim environment so it stays reliable, maintainable and usable across robotics development.
• Digital twins and asset pipelines - Keep models of robots, tooling and the wider cell physically representative as the hardware changes, with robust workflows for taking engineering geometry into USD.
• Physics that matches reality - Tune contacts, friction, joints, actuators and sensors for stable, useful behaviour, and diagnose where simulation diverges from the machine.
• ROS 2 integration - Own the interface so robotics software runs against simulated and physical systems through consistent abstractions.
• Virtual development and automated testing - Build scenarios that exercise behaviours, changes and failure conditions repeatedly, and turn them into tests that live in the robotics CI workflow.
• Sim-to-real - Find the discrepancies between simulated and real MicroFactory behaviour that actually affect engineering decisions, and close them.
• Data for what comes next - Version simulation runs as usable data, and generate synthetic training data with domain randomisation to support computer-vision development.

What good looks like

• In your first weeks, you're running the existing environment yourself, finding where it diverges from the physical MicroFactory, and forming a view on where simulation has the most leverage.
• Within your first month, another engineer can run a useful scenario without you configuring it for them.
• Within two months, there's a clear technical direction: what gets modelled, to what fidelity, and how it stays aligned as hardware changes.
• Simulation is routinely used to develop, test and make engineering decisions, not to produce demos.
• The environment keeps pace with changing hardware and software without becoming fragile or expensive to maintain.
• You know when fidelity is enough to answer the question and when physical testing is still required, and you don't gold-plate.

Who this is for

You're a senior robotics or simulation engineer who has built simulation environments for complex physical systems, and you know the gap between a simulation that looks convincing and one that is useful for engineering. You've worked where software, controls, mechanical systems, sensing and physics all interact, and you're comfortable going deep into a physics problem, debugging a ROS integration, or deciding a model isn't worth building at all.

You like owning technical infrastructure rather than contributing individual scenes. You think about architecture, maintainability and how other engineers will use what you build. You'll have real autonomy over the approach, alongside a robotics team that will use and challenge it.

If your simulation work has been about convincing visuals or research demos rather than tools other engineers depend on, or you'd rather be handed scenarios to build than decide what's worth modelling, this isn't the right role. The tooling you'd want handed to you is the tooling you'd be building.

What we're looking for

• NVIDIA Isaac Sim: significant hands-on experience building robotics simulation environments in it.
• Simulation physics: contacts, friction, joints, stiffness, damping and control parameters, and the ability to diagnose unstable or non-physical behaviour rather than tuning blindly.
• Strong Python, and fluency with Linux, Docker and Git.
• Integration with real robotics software: you've connected simulation to systems that actually run, with working knowledge of ROS 2 interfaces and workflows.
• CAD and 3D assets: engineering geometry in a robotics, manufacturing or automation context.
• Infrastructure other people use: simulation infrastructure a wider team could use and maintain, not only you.
• Judgement on fidelity: you know when a model is accurate enough, and can defend the trade-off between complexity and usefulness.
• Cross-discipline communication: you work clearly with robotics, controls and mechanical engineers, not just other simulation people.

Bonus if you have

• URDF and CAD-to-URDF/USD asset pipelines.
• Automated simulation testing, CI, or hardware/software-in-the-loop workflows.
• Industrial robots or automated manufacturing systems, and C++.
• MuJoCo or other simulation platforms, deeper PhysX knowledge, or deformable-body and compliant-joint modelling.
• ML training-data pipelines, domain randomisation, synthetic data generation or sim-to-real calibration, and cloud-based simulation infrastructure.

What we offer

• Competitive salary based on experience
• 28 bookable days + bank holidays + seasonal office closure
• Health and dental insurance via Bupa
• An 11,000 sq ft R&D and demonstration facility in E17, where you work alongside the hardware your software drives, plus time on MicroFactory sites where it runs
• Share options so we all grow together
• High autonomy, high accountability, and a team of industry experts that takes the mission seriously

Our values

• #ownyourmission - we take responsibility and execute with focus and urgency.
• #forpeopleandplanet - we strive to do the best for people and planet.
• #buildtogether - we succeed through deep collaboration with our team, partners and customers.
• #gooddesign - we centre good design in everything we do, holistically.
• #beboldberadical - we think creatively; we are ambitious, bold and radical.

This is a London role, 5 days a week in our E17 office.

You'll need the right to work in the UK. We can sponsor visas if you're a standout candidate.

We're proud that 50% of our technical team and 50% of our leadership are women. We're building a team as diverse as the communities we serve. If you see a part of this role that excites you, even if you don't meet 100% of the criteria, we encourage you to apply. Let's make housing better, together.

We are committed to creating an equitable, inclusive and diverse workplace. We intend to run an equal, fair and inclusive hiring process. If you feel there is any area we could improve, please let us know at [e-mail].