This project aims to use reinforcement learning to control an underwater robot.

You'll use a new underwater robot (ROV - remotely operated vehicle) to investigate using machine learning for control purposes. This will involve some simulations in MATLAB/Simulink, and the opportunity to apply the controllers that you develop to the BlueROV owned by the School. This is a unique opportunity to do reinforcement learning and deep learning and apply it to a real-world system.

School

Mechanical and Manufacturing Engineering

Research Area

Robotics | Control engineering | Underwater robotics | Autonomous systems | Reinforcement learning | Deep learning | Machine learning

I have weekly 30-minute meetings with my thesis and ToR students, as well as a bi-weekly group meeting where we share progress and help each other with debugging or other thorny problems.

  • Poster showing the aims and outcomes of the project
  • A control algorithm that has been tested in a simulated environment (e.g. Simulink)
  • A control algorithm that has been implemented on an underwater ROV
Senior Lecturer in Robotics and Mechatronics Will Midgley
Senior Lecturer in Robotics and Mechatronics