Research Fellow (Multi-Agent Path Planning for Autonomous Drone Operations)
Core
Develop learning-based model predictive control (MPC) algorithms for multi-agent multirotor drone navigation around vessels in maritime environments.
Role type
Research Fellow (Multi-Agent Path Planning)
Builds
Simulation-based validation frameworks for drone airspace usage, vessel protected volume estimation, and safe corridor allocation.
Domain
Maritime robotics and autonomous drone operations
Deliverable
production ML models
Required skills
multi-agent reinforcement learning, model predictive control, trajectory optimisation, Python programming, deep learning frameworks (PyTorch/TensorFlow), autonomous navigation, conflict detection, collision avoidance, robotics simulation (ROS/ROS2, Gazebo, AirSim, Unity)
Preferred skills
vessel-motion prediction, dynamic obstacle avoidance, separation assurance, traffic-aware navigation, geospatial data, maritime traffic data, Automatic Identification System data, airspace-capacity modelling, control algorithms for embodied AI systems
Responsibilities
Develop multi-agent path-planning frameworks for safe maritime operations; formulate navigation problems including state/action representation and safety constraints; develop MPC algorithms for dynamic conditions; design simulation experiments and validation studies; prepare technical reports and research publications
Seniority
Senior, hands-on IC
