I am a Ph.D. at the Dyson Robotics Laboratory under the supervision of Dr. Leutenegger and Prof. Davison. My research concerns the scalability of SLAM algorithm in space and time. We are investigating sharing a map accross multiple localisation and mapping sessions (time sequentially) as well as when multiple robots are performing SLAM in parallel: how can we find a consensus for what the map should look like? What role play dynamic objects or content that slowly moves over time? How can a map be shared and incrementally build across multiple robots?

Among my interests there is also the fusion of inertial sensors with visual information to increase the accuracy of the estimation.