We are seeking a highly motivated Principal Control Engineer to advance the state of the art in robotics behavior learning and manipulation for Boston Dynamics’ Stretch robot operating in complex warehouse environments. In this role, you will have the chance to shape the next generation of intelligent manipulation by applying modern AI and data-driven methods to improve robotic behavior planning and control.
In this role, you will work across research and production boundaries—developing learning-based approaches that can both demonstrate new capabilities and be deployed reliably in real-world operations.
Responsibilities:
Develop and evaluate learning-based solutions for robotic manipulation and behavior planning using techniques such as reinforcement learning and imitation learning.
Leverage multimodal inputs (vision, depth, language, proprioception) and modern ML architectures to improve decision making.
Collaborate closely with infrastructure engineers to build and iterate on AI/ML development systems, spanning data collection, model training, validation, and deployment on real robots.
Develop and maintain simulation systems that support large-scale data generation, model training, and performance evaluation for learning-based policies.
Collaborate with perception, planning, and control teams to integrate learned components into the broader autonomy stack.
Transition promising R&D prototypes into robust, production-quality systems deployed in the field.
Contribute to the infrastructure and tools enabling scalable AI experimentation, as well as data and model management for continuous learning and deployment.
We are looking for:
MS with 3 years of industry experience or PhD in Computer Science, Machine Learning, Robotics, or a related field
Proven experience developing in large, collaborative codebases (Python and C++).
Deep expertise and hands-on experience in one or more of the following areas:
Imitation Learning, Unsupervised Learning, or Reinforcement Learning at scale.
Designing data collection strategies, curating large-scale proprietary datasets, and developing data flywheels to drive continuous model improvement.
Learned visual representations for decision making and manipulation.
Multimodal learning, integrating vision, language, and robot states through specialized encoders.
Experience developing and using simulations to support AI/ML workflows, including data generation and model evaluation.
Demonstrated ability to bridge R&D and product development, delivering deployable technologies.
Experience in technical leadership on a cross-displine team, excellent communication and collaboration skills.
Bonus Points:
Experience with physics simulation such as MuJoCo, IsaacSim.
Background in robotic manipulation or warehouse automation.
Familiarity with on-robot perception and control integration for complex tasks.
Familiarity with ML deployment on edge or embedded systems.