What is Precision Weed Management?

Precision weed management is a catch-all term for weed management practices that are applied to different areas or to specific plants within a field, based on where problematic weeds are located. This kind of weed management is also known as site-specific weed management or targeted weed management

This is a research-scale version of the See & Spray Ultimate, a sprayer from John Deere and Blue River Technology. It can identify and spray only where the weeds are present within a crop field, an example of precision weed management in action. (Photo credit: Claudio Rubione, GROW) 

The goal of precision weed management is to increase the efficiency and effectiveness of management practices by using the right tool, at the right intensity, at the right time and only where needed.

For example, a farmer might use sensors on a spray boom to detect and spray only when weeds are present, or a drone camera might identify areas of a field where a cover crop didn’t grow well – alerting a farmer that they may need to adjust their herbicide usage or fertilizer rates to those areas.

Camera systems can also create maps of crop performance and weed populations within a field, which can help the farmer monitor the effectiveness of their management and make adjustments where necessary. Maps can also inform future management plans. As in the examples above, precision tactics can reduce the amount of herbicides and other inputs needed for crop production while improving overall weed management outcomes and crop yields compared to conventional management practices. This can result in lower production costs and/or higher revenues for farmers, leading to higher farm profitability

A camera mounted above the spray boom detects weeds in real-time on this research-scale version of John Deere’s See & Spray Ultimate. (Photo credit: Claudio Rubione, GROW)

Many forms of precision weed management use camera sensors to detect and distinguish weeds from crops. For example, robotic weeders are being developed to detect and kill weeds based on differences in size and height compared to crop plants. 

Cameras, sensors and data processing units mounted on spray booms that can identify and spray individual weeds within a growing crop are at or nearing commercialization, from companies such as John Deere, Blue River Technology, Greeneye Technology, One SmartSpray and Billberry





Dual spray booms on John Deere and Blue River Technology’s See & Spray Ultimate sprayer allows farmers to run both broadcast applications and targeted broadcast applications simultaneously. (Photo credit: Wyatt Stutzman, Virginia Tech)

These precision sprayers utilize dual tank systems, which allow applicators to run one set of chemicals through one set of nozzles for broadcast application, while another set of nozzles delivers a different load of chemicals to targeted weeds. 

Precision weed management that does not rely on active sensors typically utilizes GPS-guided tractors or robotic platforms and information about field boundaries as well as crop row locations and spacing to apply management practices to areas where crop plants are not expected to be. Examples include inter-row herbicide applications and precision cultivation practices such as band sowing and hoeing

GROW researchers are actively working on precision weed management projects, including collaborations with Precision Sustainable Agriculture (PSA), a public ag tech development organization. 


The National Ag Image Repository (NAIR) 

GROW and PSA researchers are actively building a National Ag Image Repository of weeds, cover crops and common cash crops such as corn, cotton and soybeans. 

The goal is to create a database of images collected with uniform protocols that can be used to train sensors and other computer vision models to identify weeds, cover crops and crops to the species level in real time. While such databases exist within private industry, the National Ag Image Repository will be public and open-access, with an aim of speeding the development of precision ag technologies and making it more accessible and affordable for the entire agricultural community. 



Researchers are using both robotics, such as the automated Benchbot on the left, and in-field protocols, such as the team on the right, to capture the enormous diversity of images needed to populate the National Ag Image Repository. (Photo credits: Claudio Rubione, GROW)


GROW and PSA are collaborating on the development of PlantMap3D, a low-cost camera system used to map the species, biomass, and densities of weeds, cover crops, and cash crops.

High-speed cameras mounted to tractor booms can help researchers rapidly assess and identify the biomass and density of weeds, cover crops and cash crops. (Photo credits: Claudio Rubione, GROW)

These programs use off-the-shelf cameras, specifically the OAK-D (Luxonis) multi-sensor smart camera, and can be used as a hand-held unit, mounted onto the tractor, or networked across large spray-booms. This technology is currently being tested by weed scientists and on-farm research networks across the US, with the goal of having commercial products available by 2028.

Decision Support Tools (DSTs)

GROW and PSA researchers are developing digital tools that assist growers with common farm and weed management decisions, such as what cover crop species best meet their needs or what herbicides are effective on problematic weeds in their region. 

Some of these DSTs are online and available from PSA now. See the Northeast Cover Crop Species Selector and the Cover Crop Nitrogen Calculator

Information from mapping systems such as PlantMap3D can be integrated with these tools to allow them to offer even more detailed and informed recommendations. For example, maps of cover crop biomass will eventually be imported into the Cover Crop Nitrogen Calculator to provide variable rate fertilizer recommendations based on predicted nitrogen release after cover crop termination.



Eugene Law
April Dobbs
Daniel Ginn
Matthew Kutugata
Steven Mirsky
Søren Skovsen

Emily Unglesbee 
Mark VanGessel
Michael Flessner