DrAndrew Hill

Well-designed autonomous vehicles can benefit industry by acting more consistently and reliably than human-operated vehicles, increasing efficiency and safety. Fleets of autonomous vehicles can offer further benefits by optimising their actions to reduce congestion and redundant work, lowering fossil fuel consumption and carbon emissions. All of these applications require robust planning and control algorithms, and this is the focus of Dr Andrew Hill's research.

"My work focuses particularly on using fleets of automated vehicles to perform large-scale industrial operations, such as hauling materials in mining or agriculture.

"While it's relatively easy to control a single autonomous vehicle driving on a straight road in good weather, controlling a whole fleet of autonomous vehicles that have to interact with each other in different conditions and scenarios can be much more complicated.

"Controlling a single vehicle typically means following a simple set of road rules designed for safety. Controlling a fleet requires you to optimise how these rules are followed for the fleet, deciding which vehicles should go where and in what order, constantly monitoring and re-evaluating and making predictions based on uncertain performance. It becomes even harder in the 'real world', where things rarely go to plan.

"Field robotics is the future of automation in the real world. It involves autonomous devices that operate outside the bounds of factories and production lines, instead interpreting and interacting with the world, with other machines and with people.

"While some people see the goal of robotics as 'replacing jobs with robots', that's really not the purpose. A lot of my work is about safety, such as moving people away from working in harsh desert environments or with dangerous explosives. It's also about efficiency, in that a lot of these tasks are boring and laborious, so people get distracted or cut corners and make mistakes. A person directing a robot to do these tasks still needs to make the same decisions, but the execution can be done more efficiently, safely and often with higher precision by a robotic vehicle.

"I've always been interested in making things work on their own. From Lego and Meccano to programming, I find building machines and making them do useful or interesting things a fascinating process."