Indoor flight lab clears the way for advanced aerial robotics
The indoor flight lab gave Dr Nguyen the space to build a drone and develop the algorithms to control it.
A Federation University researcher has developed software that enables drones to fly autonomously in cluttered environments, allowing them to navigate around moving obstacles and other barriers to reach their destination.
Dr Thai Binh Nguyen, who recently completed the work for his PhD, developed the smart system for drones and built a fully functional indoor flight laboratory for aero systems research at Federation’s Gippsland campus.
Unlike most autonomous systems, the algorithms developed by Dr Nguyen can be applied to lower-powered and more cost-effective devices. The algorithms have also been made open source, allowing other users to download and potentially adapt them for other platforms, such as ground vehicles or submarines. He developed a set of algorithms to enable various setup options and test multiple constraints for his thesis.
“With this system, it could be potentially used for search and rescue operations or surveillance – complex operations in very cluttered environments where there are obstacles, like a bushland setting,” Dr Nguyen said.
“I have a background in unmanned aerial vehicles (UAV) and have seen that there is a gap in the literature and in the industry in terms of the capability of drones when it comes to environments that are not straightforward. This has really motivated me to dig deeper into this area.”
Dr Nguyen says that many companies build UAVs to focus on different applications, such as defence and security, and there are often complex and differing perspectives between industry and research on UAV development.
He says that making published research open source allows other researchers and stakeholders to potentially develop various applications from the algorithms, as well as different types of robots. It also allows researchers to reproduce the algorithms to verify his results.
A key component of the project was the development of an indoor flight lab at the Gippsland campus, giving him the space to build a drone and develop the algorithms to control it.
This involved converting an empty space into a fully functional flight arena capable of supporting cutting-edge research in aerial robotics. Built from scratch on a limited budget, the arena will be available for future education and research purposes.
Dr Nguyen was also involved in the project to develop artificial intelligence-driven technology for a robot to identify koalas in gum tree plantations, alerting forestry groups to their location before any trees are harvested.
“The drone was tested in many iterations, and it was regularly optimised and developed, refining the algorithms to make it a state-of-the-art project,” he said.
“To do research, we needed a flight arena with a location and a design, and we couldn’t fly outdoors because of the potential risks, so we needed to come up with a balanced solution, being cost-effective but fully equipped for the research.”
The process involved collaborating with campus staff to identify the location and procure necessary equipment, including a large net that was installed in the lab to create an environment suitable for research and replicate real-world scenarios.
Dr Nguyen says a key point of difference with the drone that was developed was the relatively low-powered components, including the onboard photographic sensors.
“We have a minimal field-of-view camera and a limited resources platform setup, so the aim was to get the platform performing well, and for it to be competitive with other well-equipped platforms.”
“With limited resources, we were able to create an indoor flight facility in Gippsland with the fundamental functionality for state-of-the-art drone research.”