Robotic Manipulation Engineer, Lodestar

• If you're passionate about robotics and want to see your work in action, this is the role for you
• We're seeking a builder—someone who’s both research-minded and implementation-driven. You should be eager to take ownership, collaborate deeply, and see your ideas go from prototype to mission-ready
• 5+ years of experience in Machine Learning, Robotics, Reinforcement learning, or a related field (or equivalent practical experience)
• Lead the end-to-end grasping pipeline
• Understand both traditional and machine learning approaches to robot manipulation, and the tradeoffs between various techniques
• Bachelor’s or Master’s in Computer Science, Mathematics, Electrical Engineering, related field or demonstrated an equivalent level of knowledge
• Deep understanding of the fundamentals of machine learning
• Experience with machine learning frameworks (e.g. PyTorch, TensorFlow)
• Hands-on proficiency in C++ and Python
• Familiarity with robotic simulation environments (e.g. NVIDIA Omniverse, MuJoCo, PyBullet)
• Solid grasp of real-world constraints in robotic perception and manipulation
• Strong communication and collaboration skills
• You may be ineligible for this role if you do not hold citizenship of the United Kingdom, Australia, Japan, New Zealand, Switzerland, the European Union or a country that is part of NATO, or if you hold ineligible dual citizenship or nationality

Desirable

• We’re a team of roboticists, engineers, and dreamers, committed to building the future of space infrastructure, and space domain defence
• Our robotic systems are designed to perform autonomous manipulation in the harshest and most complex environments imaginable
• We’re looking for a Machine Learning Robotic Manipulation Engineer to lead development of the core ML-based grasping stack for our robotic capture system
• In this role, you’ll research, design, and implement models and pipelines that enable perception-driven grasp planning—bridging the gap between computer vision, motion planning, and real-world physical interaction with unknown objects
• You’ll work closely with Machine Vision, Robotics, Hardware, Systems, and Software teams, to take your models from the lab to orbit—from simulation to flight
• Lead development of machine learning systems for real world grasping and manipulation of unknown objects in space
• Design and train models for arbitrary grasp candidate prediction and real-time grasp quality evaluation
• Build high fidelity physics simulators to train and evaluate grasp policies
• Deploy and fine-tune grasp models to achieve real-time inference on our edge compute platform
• Collaborate with robotics, perception, hardware, software, and systems teams to ensure seamless integration
• Own performance benchmarking, real-world validation, and iteration for grasp reliability

• For security reasons background checks will be undertaken prior to any employment offers being made to an applicant. These checks will include nationality checks as it is a requirement of this position that you be eligible to access equipment and data regulated by the United States' International Traffic in Arms Regulations