Senior/Staff Software Engineer, Motion Planning

Job description

Job responsibilities

• Design and implement algorithms for decision-making and trajectory generation, ensuring safety, comfort, and reliability across a wide range of driving conditions.
• Bring novel robotics and machine learning techniques from research into a production-grade, safety-critical system.
• Reason about complex, socially-nuanced driving situations, such as negotiating merges, junctions, and interactions with other road users, under uncertainty.
• Utilize modern C++ numerical optimization frameworks (e.g., CasADi, Acados) and industry-standard tooling (git, ROS, etc.) to integrate new planning algorithms into the production stack.
• Partner closely with the ML and prediction teams to integrate, refine, and act on probabilistic and multi-future representations of how a scenario might evolve, shaping how these outputs are consumed by planning…
• Perform system-level diagnostics and tuning using in-house and open-source visualization tools (e.g., webviz).
• Collaborate cross-functionally and communicate clearly, producing system designs that are reviewed and approved at a high level before implementation begins.

Job requirements

• MS with 1-2 years of relevant experience, or PhD, in Computer Science, Robotics, or a related field.
• Strong production software development skills in C++; Python experience a plus.
• Robotics expertise in one or more of: path planning, motion planning, trajectory generation, behavior planning, optimal control.
• Strong math (geometry, analysis, probability) and physics (kinematics, dynamics) fundamentals.
• Experience designing, implementing, and refining planning algorithms on a real-world robotic or vehicle platform.
• Strong oral and written communication skills, with the ability to collaborate across teams and derive solutions from cross-functional input
• PhD + 1-2 years or MS + 3-4 years of relevant experience.
• Bonus: exposure to machine learning techniques such as deep learning, imitation learning, transformer-based models, or diffusion-based models, particularly as applied to decision-making or scenario generation.