Skip to main content
Robotic arm
Centres and institutes_

Planning, dynamics and control

Enabling robots to plan, move, act, and learn from experience
We’re creating new technologies that will enable the next generation of robots to act in the real world, make complex decisions about where to go and what to do, and to learn over time from accumulated experience.

Robotics is the science of the cyber-physical interface – the connection between the cyber-world of information and computation and the real-world of things, masses, and forces.

Our research is focused on developing new motion planning and control algorithms that enable robots to interact with the real world in challenging circumstances.

We’re investigating new technologies to ensure safe and efficient movement through the environment, from wheeled and legged ground robots, to fixed and rotary-wing aerial robots, and surface and submersible marine robots,

We’re also developing novel mechanisms and manipulation technologies to enable robot arms and hands to perform productive tasks.

Our system identification and machine learning techniques allow robots to learn to predict the future based on past experience and use these predictions for robust and reliable planning and control.

Our research

Our expert: Associate Professor Ian Manchester

Our collaborators: Professor Thomas Schön (Uppsala University, Sweden)

This project aims to make machine-learning of dynamic system models reliable, accurate, and secure. Robots and other autonomous machines use models of the real world to predict the result of their actions and make decisions, but existing methods to learn such models from data are unreliable in many cases and can be easily fooled.

The outcomes of this project will be new models and algorithms that ensure safety and increase accuracy of models learned from data. This will benefit robotics, control engineering, infrastructure automation, and other fields that demand the capability to model physical systems from limited data. It will also improve cybersecurity by making learning algorithms resilient to deliberate attacks with false data

Our expert: Associate Professor Ian Manchester

Our collaborator: Professor Jean-Jacques Slotine (Massachusetts Institute of Technology, USA)

The coming generation of robots are highly mobile and will interact significantly with their environment, each other, and human collaborators. However, this leads to highly coupled nonlinear dynamical behaviour, and achieving accurate and reliable control of these systems is pushing current control theory to breaking point.

In this project, we will develop a new approach to control of nonlinear systems based on contraction theory and convex optimisation, extending the power of optimisation-based control from linear to non-linear systems. The project will lead to new theoretical developments, constructive algorithms and software, and experimental demonstrations on a range of platforms including bipedal walking robots and underwater robots.