Robotics Learning Engineer – Humanoid Robots, Real-World ML
Posted Updated
Description
Latent Edge Ltd in London is seeking a Robotics Learning Engineer to develop learning systems for humanoid robots using deep learning and reinforcement learning. You will deploy models from training to real hardware, optimize data, and collaborate with simulation and controls engineers.
Join a fast-moving team working at the intersection of robotics, AI, and hardware to advance industrial humanoid robotics.
Develop and fine-tune robot manipulation policies using behaviour cloning and reinforcement learning. Train, evaluate and deploy ML models onto real robotic hardware. Improve training datasets through preprocessing, curation and quality optimisation. Work with simulation engineers to develop RL environments and improve sim-to-real transfer. Collaborate with data collection teams to improve training data quality, diversity and coverage. Develop new observation and action spaces to support advanced robotic manipulation. Work with teleoperation and controls engineers to improve robot movement, responsiveness and smoothness. Investigate hardware and control limitations and contribute to future robot design improvements. Run experiments, analyse results and identify ways to improve robotic performance. 3+ years of robotics systems experience in industry or research. Hands-on work deploying software on real robots. Experience with deep learning and neural network training. Strong Python skills and PyTorch/JAX familiarity. Knowledge of robotic control, teleoperation or motion planning. Experience with ML datasets, training and evaluation. Software engineering skills including debugging and testing. Ability to analyze results and solve complex problems. Private healthcare Pension contributions Free daily breakfast Office snacks Equity participationSkills
Sign in to read and join the discussion.