Drones aim to identify potato virus Y plant by plant
Montana State student Shehzore Hashmi is combining drone color and multispectral images with machine learning to identify potato virus Y in individual plants and make field scouting more efficient.
- Entered
- Plot
- 1.6 ha
- Record
- 620 words
- Walk-through
- 3 min
DrillSprayHarvestPlanning window

Agronomist’s notes
- Shehzore Hashmi is a second-year doctoral student at Montana State University.
- He studies electrical and computer engineering in the Norm Asbjornson College of Engineering.
- His team combines color photography, multispectral imagery, machine learning and computer vision.
- The method targets individual potato plants showing potato virus Y symptoms.
- Hashmi still needs his drone pilot’s license and is considering careers in academia or industry.
By James Calloway
Shehzore Hashmi still needs his drone pilot’s license to pursue Montana State University research that identifies potato virus Y in individual potato plants. He is a second-year doctoral student in electrical and computer engineering at MSU’s Norm Asbjornson College of Engineering.
What is Hashmi trying to detect?
Potato virus Y (PVY) infects potatoes and other crops. Disease associated with the virus can reduce plant health, crop quality and yield, giving Hashmi’s engineering project a clear agricultural target.
His work sits within an interdisciplinary team at Montana State University. The team combines standard color photography and multispectral imagery with machine learning and computer vision to locate individual potato plants showing PVY symptoms.
The intended result is more efficient identification than traditional field scouting provides. Hashmi is studying how those technologies work together, rather than treating drone photography as a complete diagnostic system on its own.
How does the imaging system work?
Standard color photography records the scene visible to the camera operator. Multispectral imagery captures wavelengths of light that the human eye cannot see, adding information about crop conditions that may not appear in an ordinary photograph.
Machine learning provides methods for recognizing patterns in data. Computer vision interprets image content, allowing the combined images to be assessed for features associated with symptomatic potato plants.
The drone serves as the aerial imaging platform for collecting both image types. The analytical stages then combine the visible and multispectral information to identify plants showing signs of potato virus Y.
Why does plant-level detection matter?
A grower facing a field-level disease problem still needs to know which individual plants show symptoms. Plant-level identification gives a more precise basis for deciding where inspectors should focus their attention.
The practical comparison is direct: traditional field scouting relies on inspecting crop plants in the field, while the proposed workflow reviews aerial images and identifies individual candidates for closer checks.
That distinction could reduce the time needed to locate symptomatic plants. However, no measured time saving, detection rate or acreage result has been published, so the size of the advantage remains undefined.
Identification also has limits. The method finds plants showing symptoms associated with PVY, but the available information does not establish that it prescribes treatment or replaces laboratory confirmation.
What does the project need to show next?
Hashmi’s pilot’s license remains a practical prerequisite for his own drone flying. Until licensed flights and subsequent scouting comparisons are available, the work remains university research rather than a service growers can buy.
Several commercial and performance details are still missing:
- An accuracy rate for identifying symptomatic potato plants.
- Hectares covered by each drone survey or flight.
- Flight time and labor required compared with conventional scouting.
- A drone model, commercial price or farm deployment date.
Those figures would determine whether the workflow changes a routine scouting decision or merely demonstrates the technical ability to classify plants from imagery.
Hashmi is considering a career in academia or industry. That choice leaves the eventual route open, but no product timetable or commercial partnership has been announced.
What should farmers watch?\n
The next evidence should come from licensed drone work and direct validation against traditional field scouting. The relevant comparison is whether the system identifies individual symptomatic potato plants more efficiently, and how often those identifications prove correct.
Researchers should also report the imagery system, field conditions and scouting protocol used in any comparison. Without those details, growers cannot judge whether the approach suits their own fields or management practices.
For now, the project has no stated launch date or farm availability; farmers should watch for Hashmi obtaining his license and for the next reported comparison with conventional field scouting.
via High Plains Journal (Source)
More from James Calloway
Show full bio
Correspondent covering regulation, policy and their impact on production farming and field operations at Arable Wire.
201 articles
Nearby records
- AV 15Croptimal launches €95,000 mounted Croptiscan 3000 for seed potatoesOctober 10, 2026
- EZ 41Three crop scouting apps South African growers can download todayNovember 8, 2026
- HA 69CropX buys SCIO for NIR crop quality data — eighth deal since 2020October 10, 2026
- NS 50NASA partners with farmers on satellite crop monitoringOctober 10, 2026
- UU 63John Deere Bets on AI to Drive 'Decision Agriculture' ShiftOctober 1, 2026