Robotics in Agriculture Gets an Assist from USDA-ARS

Agriculture is poised to benefit greatly from the rising revolution of robotics and artificial intelligence – if we can actually get those tools into the hands of farmers. 

That’s a big if, but USDA’s Agricultural Research Service (ARS) and land-grant scientists are well positioned to make it a reality, says Dr. Steven Mirsky, a USDA-ARS research ecologist, as well as co-founder of the GROW and Precision Sustainable Agriculture (PSA) networks. 

The key is to build large, open-source image recognition pipelines that train machines and computers to rapidly identify and target individual plants and pests, Mirsky says. While these types of resources exist behind the walls of private industry right now, Mirsky and his research teams are working to make them more widely and publicly available, with projects such as the Ag Image Repository and PlantMap3D

“Our aim with robotics in agriculture is to help farmers make increasingly complex decisions, improve their quality of life and empower them to use more sustainable farming strategies.”

Dr. Steven Mirsky, USDA ARS

The aim is to speed up the development of robotics in agriculture, which Mirsky believes will actually help address the trend of shrinking human participation in agriculture and food production. 

“We hear all the time that people are worried about robots stealing jobs and pushing farmers out of the industry, but folks, that exodus has already happened,” Mirsky explains. “Since the Industrial Revolution, we have shifted from 90% of the population living on farms to just 1%. That’s a massive loss of human intelligence – fewer people with eyes on the ground.” 

By harnessing the power of robotic systems, which are trained via thousands of plant images and increasingly move autonomously through fields, we can try to reverse that trend, Mirsky says. “The digital agriculture revolution is bringing many eyes – and as a result, intelligence – back to the farmers, through much-needed low-cost, sensing technology and robots,” Mirsky says. 

He hopes that USDA’s leadership on open-source, AI-training databases in agriculture can speed up research and development on precision weed management technology (such as target sprayers or drones) and lower their price tags, as more companies and systems enter the market. 

The end goal is for farmers to use this new technology to fine-tune their farm management systems, spraying pesticides or applying fertilizer and water precisely (and only) where needed – benefiting both their bottom line and the climate. 

“Our aim with robotics in agriculture is to help farmers make increasingly complex decisions, improve their quality of life and empower them to use more sustainable farming strategies,” Mirsky says. 

Learn more about this topic on GROW’s Precision Weed Management page, as well as the Precision Sustainable Agriculture website.


Article and feature photo graphic by Emily Unglesbee, GROW; video by USDA ARS; header and feature photos by PSA