This project explores the implementation of Model Predictive Control (MPC) for autonomous navigation and motion control in mobile robots. MPC is a powerful control strategy that optimises future control actions based on a predictive model of the robot's dynamics and constraints. The student will select one robot platform (listed below), develop an MPC framework tailored to its mobility and sensing capabilities. The project will involve modelling the robot’s kinematics, designing the MPC controller, and validating its performance in simulated and optionally, real-world navigation tasks. Emphasis will be placed on trajectory tracking and real-time responsiveness. Implementation, deployment and demo should be done using the Robotics Operating System (ROS2).

Robots:

Reachy (Pollen Robotics): https://www.pollen-robotics.com/reachy/

Stretch (Hello Robot): https://teal-blue-zpt3.squarespace.com/stretch-2

Unitree A1: https://unitreerobotics.net/robotdog/unitree-a1/

Formula Student AI racing car: For this option you should be involved or get involved with Cardiff Autonomous Racing (CAR) https://www.imeche.org/events/formula-student/team-information/fs-ai https://github.com/FS-AI/FS-AI_ADS-DV_CAD