Robotics Engineer; planning & behaviour

Department

Behaviour

Job Type

Full-time

Location

Cambridge

Modality

Hybrid

Start Date

Empty

We are looking for a Robotics Engineer to join our Behaviour Team . You will build the core planning and action-execution capabilities of the Mastermind system. This role is responsible for the full stack of robotic behaviour, from high-level strategic planning down to the interaction with low-level motion control sub-processes. You will architect and implement both a task manager, which determines the multi-agent strategy to achieve a mission , and a behaviour manager, which translates those strategies into collision-free physical motion for our fleet of robots.

Note:

you will not need to touch or implement low level proportional or PID controllers.

• Strategic Task Planning:
  Design and implement multi-agent planning algorithms that decompose complex mission goals into a sequence of high-level tasks for the robotic team.
• Motion Planning & Control:
  Design and implement software that translates high-level behavioral requests into executable, low-level commands for various robotic platforms (e.g., vehicles running PX4).
• Heterogeneous Planner Integration:
  Integrate and manage a library of diverse motion planners for our heterogeneous team of robots, including navigation stacks (e.g., TEB planner) and manipulation planners (e.g., Move It).
• Action Abstraction:
  Define a clear hierarchical interface between high-level strategic tasks and the low-level robotic actions required to execute them (e.g. using PDDL).
• Feasibility and Validation:
  Implement the geometric reasoning and feasibility checks to verify that actions are physically possible, providing critical feedback to the system to trigger adaptive responses.

• MSc or PhD in Robotics, Computer Science, or a related engineering field.
• A strong, demonstrable background in robotics planning, spanning both high-level task planning and low-level motion planning, control, and kinematics.
• Experience with simulation tools and workflows (Isaac Lab preferably, Mu Jo Co , SAPIEN, Gazebo are also relevant).
• Reinforcement learning and automated planning (e.g. using PDDL).
• Extensive, hands-on experience with ROS 2 and standard planning frameworks within its ecosystem (e.g., Nav2, Move It).
• Familiarity with collision checking, obstacle avoidance techniques, and geometric data structures used in planning.
• Excellent, production-quality programming skills in C++ and Python.

• Practical experience with symbolic AI planners or formal planning languages like PDDL.
• Experience in multi-agent planning, task allocation, or auction-based mechanisms.
• A background in trajectory optimization or optimal control theory.
• Experience developing custom controllers and planners for real-world robotic hardware.