CSIRO's Drone Fleet Transforms Environmental Research Across Australia

Amanda Meys has never piloted a plane, yet she oversees nearly 100 remote pilots across Australia's national science agency. Her title is chief remote pilot at CSIRO, and the aircraft under her command are not conventional—they are drones, a fleet that has grown from 25 to 30 units in 2020 to around 160 today. The expansion mirrors a broader shift in how Australia's research community works. Civil Aviation Safety Authority records show the number of licensed remote pilots has climbed from 3,300 in 2016 to 16,500 by 2020, and now sits near 45,000, not counting recreational users. In less than a decade, drones have become ordinary tools for environmental science.

Meys came to the role in 2019 with no drone experience but a background as an avionics technician, maintaining aircraft electrical and radar systems. She now manages fleet operations, trains certified pilots, ensures regulatory compliance, and approves flight plans. The drones themselves range from 2 kilograms to over 100 kilograms and can operate at night. What has transformed the technology is not the aircraft but what they carry—the software, cameras, and sensors have advanced dramatically. "The drones themselves haven't changed much, but the software and the technology and the camera systems have certainly advanced," Meys says. For her, aviation has been a lifelong passion, and watching the field grow has been its own reward.

In the forests of eastern Australia, thermal imaging drones are changing how scientists count koalas. The animals are notoriously difficult to locate from the ground. Researchers like CSIRO field ecologist Eric Vanderduys rely on acoustic recorders to detect the bellowing of males along the coast from Victoria through Queensland, but sound tells you presence, not numbers. When his team began using drones in 2022, two advantages emerged immediately. Drones can fly above the canopy where humans cannot reach, and equipped with thermal cameras, they detect heat signatures invisible to the naked eye. On overcast days or after rain, when a koala's body warmth stands out against cool vegetation, the thermal image reveals what ground surveys miss. The drones fly in straight lines called transects, like a lawn mower cutting a field, which means the same route can be repeated years later at identical height and resolution. In ten years, scientists can compare the same transect and track population changes with precision. This repeatability is how population models inform species protection decisions.

Farther inland, LiDAR technology is reshaping how researchers map vast landscapes. Dr. Shaun Levick, a senior principal research scientist, once trudged through remote terrain with heavy equipment or contracted expensive light aircraft. Now he deploys drones equipped with LiDAR sensors—Light Detection and Ranging systems that use rapid laser pulses to create three-dimensional maps. The breakthrough came around 2021 with the introduction of the Riegl VUX-120 sensor, which delivered ground-quality measurements from the air. What once took three hours to map one hectare now takes 30 minutes to cover 100 hectares. The sensor captures over 1,000 points per square meter, allowing measurement of individual tree height, crown diameter, and stem diameter. Levick has used this precision to estimate populations of rare species like red cabbage palms in Palm Valley within Finke Gorge National Park and Acacia Peuce, one of Australia's slowest-growing trees, in the Matt Clark Conservation Area. The data also enables biomass and carbon estimates, with applications in carbon accounting. In a single fieldtrip, researchers can collect data from up to 1,000 hectares, which then trains satellite-based imagery models and provides ground calibration for NASA-ISRO's NISAR mission and the European Space Agency's BIOMASS mission.

Coastal research presents different challenges. Dr. Paul Branson, a coastal research scientist in Western Australia, studies oceanography to understand how currents and waves shape coastal ecosystems. Drones have limited flight time, and wind is the limiting factor—the smaller the drone, the less wind it tolerates. Branson needs wind because it drives the currents and wave motions he studies, but too much wind makes drones uncontrollable. Around 10 knots is the optimal window. Equipped with cameras sensitive to different wavelengths of light, drones can observe wave signatures and seafloor features that satellites cannot target on demand. They can monitor disturbances from dredging projects, harmful algal blooms, and geomorphological features that erode over decades. The ocean is never static, and coastlines evolve across multiple time scales, making planning decisions difficult without this kind of detailed, deployable observation.

In agriculture, drones have already become standard equipment. CSIRO cotton researchers use them for crop management, applying herbicides and pesticides to specific areas and reducing the need for workers in chemical suits during summer growing seasons. Dr. Warren Conaty's team initially contracted drone services but quickly recognized the potential of collecting multiple data streams simultaneously. They built an internal pilot team and equipped drones with multispectral cameras that capture a wider range of wavelengths than human eyes can see. These signatures characterize the environment and measure the performance of individual cotton lines. In a breeding program, what you observe in a plant—its phenotype—results from both its genetics and its environment. By capturing how the environment interacts with each plant through multispectral signatures, researchers can improve the accuracy of algorithms that identify superior breeding lines. The team also uses hyperspectral imagery to identify disease pressure and genetic resistance sources, and LiDAR cameras to measure growth characteristics early and late in the season. The goal is finer-scale resolution on how plants grow in relation to their environment, enabling better identification of breeding lines with commercially valuable traits.

Meys reflects on the growth she has witnessed. When she started, only a handful of people worked with drones at CSIRO. Now the applications span wildlife monitoring, biodiversity mapping, coastal modeling, and crop improvement. "We have just scratched the surface," she says. The technology continues to mature, and as sensors and software capabilities advance, new uses will emerge. What began as a niche tool has become essential infrastructure for understanding and managing Australia's environment and agriculture.