A PhD candidate from the Queensland University of Technology has been propelled onto the international list of 2019 Women to Watch in UAS.
According to a recent press release, she was added to the list because of her innovative use of advanced detection techniques to count koala populations.
Improving general wildlife research and management
The student is part of a team that is combining machine learning and thermal imaging with drones to count wildlife in a way that is faster, more accurate, and less invasive than traditional methods.
These impressive results create a greater implication for general wildlife research and management in complex environments around the world.
Conducting research as part of the University’s Quantitative Applied Spatial Ecology (QASE) lab has been a rewarding experience for the student as she enjoys working on a project that has very practical applications.
QASE works to protect threatened koala populations using drones and artificial intelligence (AI).
The questions being researched on revolve around wildlife management, biodiversity, invasive species and food security.
As ecologists, they use statistical and mathematical approaches in a spatial framework to solve some of the critical problems facing the environment.
The group uses state of the art technologies to monitor the environment.
In addition, they collaborate with a range of organisations including federal and state governments, conservationists, farmers and other universities to develop highly innovative solutions.
Other works
Examples of their work include:
- Monitoring koala populations with Unmanned Aerial Vehicles (UAVs)
- Using acoustics to assess the flow on impacts of Myrtle Rust in Australian forests
- Assessing crop disease status with hyper spectral imagery
- Developing improved surveillance and detection methodologies that could be used anywhere
The PhD candidate saw an opportunity to apply maths, statistics and new technologies to current ecological problems.
Drones as a koala-counting technique
Being able to help address this problem on koala counting and to work with scientists, with experiences very different from her own, helped her gain new skills and work in an emerging field.
It gave her the chance to be part of a really exciting and cutting-edge transdisciplinary project, scholarship opportunities, great mentorship and the ability to work and study close to family.
Their drones have an average accuracy of 86% for spotting wildlife, compared to the 70% accuracy of an expert human spotter.
However, this does not mean drones should or could replace human spotters.
Each method has advantages and limitations based on the environment and terrain and the team has been working on ways to combine the two methods to achieve the best outcomes.
Her research project has also attracted international attention. She has also been invited to chair a panel in the upcoming World of Drones Congress in Brisbane.