The aim of this work is to improve robotic navigation techniques for use in entirely natural environments, such as forests. It is undertaken in collaboration with neuroscientists tracking animals for behavioural experiments and geoscientists performing large scale forest monitoring.
There is a great deal of literature in robotics on visual Simultaneous Localisation and Mapping (SLAM), but the common environments for their use are well-organised and structured by human beings, so the solutions produced do not necessarily perform well in complex, dynamic natural environments. We are investigating the limitations of the state of the art in this setting and developing improvements to address them.
Research interests: Vision, Navigation, Localisation, Mapping, Reconstruction, Bio-inspired Robotics