Software Engineer, Simulation (Digital Twins)

About the role

You will be the first engineer on the Simulation team, responsible for creating repeatable digital twin pipelines from lidar maps. Your work will enable the training of multi-robot patrol policies and ensure that learned skills transfer effectively from simulation to real-world robotic deployments. You will collaborate with the Skills, Orchestration, and Reliability teams to build scalable, high-fidelity simulation environments.

Responsibilities

• Develop a pipeline to convert lidar maps into simulation-ready digital twins including geometry, semantics, materials, and sensors.
• Create reinforcement learning environments for skills such as multi-robot patrol and search, ensuring randomization to prevent overfitting.
• Generate rare event scenarios including intrusions, sabotage, weather conditions, and sensor failures.
• Develop realistic sensor simulations for RGB, thermal, and lidar data.
• Measure and minimize the sim-to-real gap on actual robot deployments.
• Scale simulation infrastructure to run many parallel environments on GPUs for fast training cycles.

Required skills

• Robotics simulation
• RL training environments
• Isaac Sim
• Isaac Lab
• Python
• 3D data
• Point clouds
• Meshes
• USD
• GPU-parallel simulation
• Reinforcement learning

Nice to have

• C++
• MuJoCo
• Gazebo
• Legged robots
• Photogrammetry
• Gaussian splatting
• ROS 2
• Thermal sensor modelling
• Lidar sensor modelling

Benefits

• Equity package
• Career growth opportunities

About the Company

Laelaps AI is a robotics company building intelligent software for physical security in the real world. Based in Zurich, the company is backed by visionary investors and led by a founding team with expertise in AI, Robotics, and Physics.