Beyond the Visible: How APPN Is Advancing Field Phenomics through Drone Operations
- 1 hour ago
- 5 min read
At the University of Sydney, Dr William “Tam” Salter and the Australian Plant Phenomics Network are combining hyperspectral, LiDAR, and thermal sensing with the Inspired Flight IF1200 to capture objective crop data at scale.
Modern crop research can involve thousands of individual field plots. Each plot may respond differently to heat, water availability, soil conditions, management practices, pests, and disease. Understanding those differences requires researchers to collect detailed measurements across an enormous area, often during narrow periods in a crop’s development.
For the Australian Plant Phenomics Network, or APPN, this challenge represents what Dr William “Tam” Salter calls the measurement bottleneck.
“The biggest problem we’re trying to solve is the measurement bottleneck,” Tam said. “Modern crop research can involve thousands of plots, but traditional assessments are often slow, labour intensive, and subjective.”
APPN is Australia’s national plant phenotyping network. Supported by the Australian Government’s National Collaborative Research Infrastructure Strategy, the Grains Research and Development Corporation, and partner organizations, the network provides researchers and industry with access to advanced sensing technologies, field infrastructure, and data science capabilities.
Tam is a crop physiologist and Director of APPN’s University of Sydney node in Narrabri, New South Wales. His team supports academic and industry researchers with field phenotyping, remote sensing, and data analysis capabilities designed to better understand how plants perform under real agricultural conditions.
Measuring Crops Where They Grow
Controlled environments remain essential for isolating and understanding specific biological processes. But crops are ultimately grown outside, where many environmental and management factors interact at the same time.
Field phenomics allows researchers to study those interactions directly. Rather than relying only on observations made in a laboratory or greenhouse, researchers can collect measurements from plants as they respond to the conditions they will encounter in production.
This helps increase confidence that findings will translate into meaningful outcomes on the farm. It can also support the development of improved crop varieties, more efficient use of resources, better farm management practices, and greater resilience to changing environmental conditions.
Historically, many field traits have been assessed visually or measured manually. Those methods can provide valuable information, but they are difficult to apply consistently across hundreds or thousands of plots.
High-throughput sensing changes that equation. It allows researchers to collect objective and repeatable measurements at a scale that would not be practical by hand.
The goal is not simply to collect more data. It is to reveal meaningful plant characteristics that traditional assessments may miss.
Seeing What the Human Eye Cannot in Drone Operations
The University of Sydney APPN node uses multiple sensing technologies to examine different aspects of crop performance.
Hyperspectral imaging captures information across many narrow bands of the electromagnetic spectrum. Through systems such as the GRYFN GOBI and CALVIS coaligned sensor payloads, researchers can quantify subtle changes in plant physiology that may occur before visible symptoms appear.
That capability could help researchers detect how plants are responding to stress while there is still little or no outward sign of a problem.
“Changes in canopy temperature, photosynthetic activity, water use, and growth patterns often occur well before visible symptoms develop,” Tam explained.
“Multi-sensor phenotyping gives us an opportunity to identify these responses early and better understand the mechanisms that underpin stress tolerance.”
LiDAR adds another layer of information. While hyperspectral data can provide insight into plant chemistry, function, and physiology, LiDAR produces detailed measurements of crop structure and architecture.
Thermal imagery provides an additional perspective by helping researchers examine plant temperature, stress, and water use. The team has also begun deploying an ITRES microTABI640 cryo-cooled thermal line scanner using the IF1200.
When these datasets are collected together and precisely aligned, researchers gain a more complete picture of crop performance. Spectral data can indicate what is happening physiologically, LiDAR can show how the plant is structured, and thermal imagery can reveal patterns associated with temperature and water use.
No individual sensor can answer every research question. The value comes from combining complementary measurements and understanding how they relate.
Turning Measurements Into Comparable Data
Collecting information is only part of the challenge. Hyperspectral, LiDAR, and thermal systems generate large and complex datasets, and those measurements must be reliable before researchers can draw conclusions from them.
APPN is developing standardized workflows, calibration procedures, data pipelines, and analytical methods that allow information to be compared across different sites, seasons, operators, and sensor platforms.
This work is especially important for a national network. Researchers need confidence that differences within a dataset reflect real biological variation, rather than differences in how, when, or where the information was collected.
APPN’s Aerial Phenomics Excellence, or APEx, program is working toward internationally accepted protocols for UAV-based field phenotyping. These protocols consider factors such as flight altitude, speed, viewing geometry, calibration procedures, and environmental conditions.
Reliable Data Starts Before Takeoff
A typical aerial phenotyping campaign involves much more than flying an aircraft over a field.
Before arriving on site, the APPN team conducts site inspections, monitors weather, develops flight plans, and addresses applicable airspace requirements. Once in the field, the team deploys spectral calibration panels, validation targets, and ground control points.
Although the hyperspectral systems do not strictly require ground control points, the additional references provide redundancy and help the team evaluate both spectral and spatial data quality.
“In scientific applications, reliable data starts long before takeoff.”
Weather remains one of the most significant challenges. Changes in illumination, cloud cover, wind, temperature, and atmospheric conditions can influence the information collected, particularly during hyperspectral operations.
When researchers are comparing hundreds or thousands of plots at a high level of precision, even small inconsistencies matter. Environmental monitoring, calibration, and flight planning are therefore fundamental parts of the scientific workflow.
As Tam noted, the aircraft is only one part of the equation. The objective is to build a complete process capable of producing scientifically defensible data under changing field conditions.
Carrying Advanced Sensing Into the Field
The aircraft still plays a critical role within that process.
Small variations in platform stability can affect sensor performance and data quality. For advanced hyperspectral, LiDAR, and thermal systems, APPN needed an aircraft with sufficient payload capacity, endurance, stability, and operational reliability.
The sensors represent significant investments, but the information they collect can be even more valuable. Once a crop passes a critical growth stage or environmental event, the opportunity to recreate that exact dataset may be gone.
Research campaigns also operate within limited weather windows. Unexpected downtime can mean missing the period when a specific plant response needs to be measured.
“For us, trust comes down to reliability,” Tam said. “If the platform performs consistently, our researchers can focus on the science rather than worrying about the aircraft.”
The University of Sydney APPN node is building a two-aircraft IF1200 capability, allowing the team to support coaligned hyperspectral sensing and expand its thermal phenotyping work. According to Tam, the IF1200 provides the payload capacity and stability required to carry these advanced systems while supporting consistent data collection across research campaigns.
Access to technical support was also an important consideration. Research teams often have limited opportunities to gather data, making responsive assistance and minimizing downtime essential parts of maintaining an operational phenotyping program.
From Better Measurements to Better Decisions
Advanced field phenomics is helping researchers move beyond isolated observations and toward repeatable, quantitative measurements of crop performance.
By integrating hyperspectral, LiDAR, and thermal information, APPN can help researchers investigate how plants grow, respond to stress, use water, and adapt to their environment. Standardized collection methods and shared data practices then make it possible to compare those findings across projects, locations, and seasons.
The result is not simply a larger dataset. It is a stronger foundation for plant science, breeding, agronomy, and agricultural decision-making.
“The platform is every bit as important as the sensor itself,” Tam said.
“Together they enable the collection of robust, research-grade data that supports better science, stronger collaboration, and ultimately greater impact for Australian agriculture.”
Read more from APPN: https://www.plantphenomics.org.au/strengthening-a-national-standard-for-uav-based-field-phenomics
