Hi! I am a Postdoc in the Robotic Decision Making Laboratory at Oregon State University, advised by Prof. Geoffrey Hollinger. My research interests include developing planning algorithms for multi-robot teams performing coordinated perception tasks, with a particular emphasis on decentralised algorithms and probabilistic reasoning. Applications include marine environmental monitoring, precision agriculture, defence, and subterranean exploration. My current research at OSU is mostly within two main projects:
I completed my PhD at the Australian Centre for Field Robotics at The University of Sydney, advised by A/Prof. Robert Fitch. My PhD thesis was titled “Planning Algorithms for Multi-Robot Active Perception”. Probably the most significant contribution of my thesis is the decentralised Monte Carlo tree search algorithm (Dec-MCTS), which generalises standard MCTS (UCT) for decentralised multi-agent planning problems and maintains similar convergence rate guarantees to UCT. You can read more about Dec-MCTS in our IJRR paper.
Previously, I worked on projects involving perception for legged robots, planning for marine robotics operations, and human-robot interaction. Most of this work was performed while I was at the CSIRO Robotics and Autonomous Systems Group with Dr. Peyman Moghadam and Dr. Navinda Kottege, and the DST Group Maritime Division with Dr. Stuart Anstee. My Honours thesis at Monash University was titled “Terrain Classification using a Hexapod Robot” and advised by A/Prof. Lindsay Kleeman, Dr. Peyman Moghadam and Dr. Navinda Kottege.
You may find my CV here. You can find out more about my research below!
This page is a work in progress…
Ph.D. in Robotics, 2019
The University of Sydney
B.E. (Hons I) in Electrical and Computer Systems Engineering, 2014
B.Sc. in Computer Science and Mathematics, 2014
A fundamental task of robotic systems is to use on-board sensors and perception algorithms to understand high-level semantic properties of an environment. These semantic properties may include a map of the environment, the presence, pose and class of objects, the behaviour of other agents, or the parameters of a dynamic field. The performance of perception algorithms can be greatly improved by planning the motion of the robots to obtain high-value observations. This motivates the problem of active perception, where the goal is to plan the observation viewpoints for a team of robots while considering both the motion constraints and the perception objectives of the task at hand.
To this end, my main body of research has focussed on developing planning algorithms for multi-robot active perception, such as decentralised Monte Carlo tree search, self-organising maps for active perception, and spatiotemporal optimal stopping. The proposed algorithms aim to address various important challenges, including: online and anytime planning, decentralised coordination, long planning horizons, unreliable communication, predicting plans of other agents, and exploiting characteristics of specific perception models.
These ideas have been or are currently being applied to a variety of scenarios, such as environmental monitoring, subterranean exploration, precision agriculture, marine operations, planetary exploration, active object recognition, and general task allocation.
A powerful new algorithm for decentralised multi-robot coordination with analytical guarantees.
Specialised learning algorithms for path planning over continuous goal regions.
Dec-MCTS introduction at WAFR 2016
Planning-aware communication teaser video for ICRA 2018
DARPA Subterranean (SubT) challenge tunnel circuit (Aug 2019) teaser
A quick introduction to DARPA SubT team explorer!
Terrain classification using a hexapod robot (Honours thesis, 2013)