This Autonomous Drone Deployment System Could Boost Crop Yields by Scaring Birds Away

Using a computer vision system to track pest birds then deploying an autonomous drone to scare them away could cut damage in half.

A team of researchers at Washington State University has turned to technology in order to keep agricultural fields free of damage from birds — by automatically deploying drones to chase them away.

"Growers don’t really have a good tool they can rely on for deterring pest birds at an affordable price," claims Manoj Karkee, associate professor of WSU's Department of Biological Systems Engineering and corresponding author of the study revealing the drone-based approach, which could save farmers millions of dollars in damages every year. "With further refinement and industry partnerships, this system could work."

The technology is relatively simple: A camera system monitors the field, using a computer vision algorithm to detect and monitor birds as they fly in and out of the area. Initially, the system simply counted the birds; a second trial proved that customized ultra-compact drones could be deployed to specific areas in response to said detection.

It's not ready for prime time quite yet, however. The team admits that work still needs to be done on testing at scale and over a long period of time. "Birds are really clever," says Karkee. "They often find ways around deterrents. We don’t want a system that only lasts for a few months or years before they stop being scared off.

"We could make drones look like predators, or have reflective propellers that are really shiny. All of these working together would likely keep birds away from those vineyards and fields. We need to research that over multiple years to make sure."

The autonomous drone deployment project follows the team's earlier studies on bird deterrent systems, which showed that manually-operated drones flying randomly could reduce the bird count in vineyards fourfold and that damage to fruit crops was cut in half as a result — while the same drones can also be used to track and assess damaged areas, alerting growers where intervention might be required and estimating yield ahead of harvest.

The team's work has been published under closed-access terms in the journal Computers and Electronics in Agriculture; Karkee says that the team plans to meat with growers and technology companies as it works towards a commercial implementation of its system. "It takes time," he admits. "But the results so far are exciting. We’re looking forward to doing more work on this project."

Gareth Halfacree
Freelance journalist, technical author, hacker, tinkerer, erstwhile sysadmin. For hire:
Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles