Targeted spray technology is revolutionizing the way crops are managed. Manufacturers tout up to 90% input savings. While there are systems that use light to identify weeds, the technology behind each brand’s offering is similar.
Cameras capture images, which are analyzed by on-board computers in real time. In fractions of a second, artificial intelligence and predictive modeling algorithms differentiate between crop and weeds, and direct nozzles to open and close.
Their precision is measured in inches. Read on for a rundown of what’s available on the market today.
John Deere’s See & Spray. The targeted application system leverages artificial intelligence to target individual weeds in real time.
“We put 36 cameras across the boom,” says Kyle Barry, manager of precision upgrade marketing at John Deere. “It’s able to identify the weeds and differentiate them from the crop.”
Several options make See & Spray available either as a factory-installed unit, or a precision upgrade on existing sprayers and select new John Deere and Hagie Sprayers. See & Spray’s three different systems — Premium, Ultimate and Select — can target weeds in fallow fields, and in corn, soybeans and cotton crops.
Greeneye Technology. Its AI-driven and retrofittable targeted spray technology uses sensors to distinguish between weeds and crop, at different growth stages and under varying conditions. It’s effective at speeds up to 15 mph.
The system works in both green-on-green and green-on-brown environments, and can identify plant species to combat herbicide-resistant weeds. It works in various row crops and was recently expanded to cotton.
“Weed control is particularly resource-intensive for cotton producers who face the added challenge of frequent spraying,” says Nadav Bocher, CEO of Greeneye.
The solution integrates with any commercial sprayer, regardless of brand, model or size.
One Smart Spray. The targeted spray system is designed in collaboration by BASF and Bosch. Camera- and lighting unit-equipped modules attach to the sprayer’s boom, targeting weeds pre- and postemergence. The retrofittable system works with corn, soybeans, cotton, sunflower, canola and sugarbeet crops.
“During application, cameras scan the field in front of the sprayer’s boom as it runs, detecting weeds in milliseconds,” reads a brand description about the system. “In combination with the system’s agronomic intelligence, each nozzle is activated only where needed.”
Weed-It’s Quadro. The precision spraying solution can be retrofitted onto ground gliders, trailed sprayers, and self-propelled or robotic sprayers. The system, which is designed for green on brown, can detect 1-centimeter weeds at 15 mph. Instead of computer vision and image processing, the Quadro system relies on fluorescence technology to identify weeds.
Sensors shine light from the spray boom onto the plants. Most of the light is absorbed by the plant and used for photosynthesis, but a small portion reacts to living chlorophyll in the plants and emits near-infrared light. The sensors measure this wavelength to distinguish plants from other substances, like dead plants or soil.
AmaSpot by Amazone. The sensor-controlled nozzle system is for green-on-brown weed detection. Like the Quadro, AmaZone uses fluorescence sensors to target living plants in green-on-brown environments. The sensors scan the ground in front of the sprayer boom and detect the areas where weeds or volunteer grains are present. Spot nozzles target individual weeds.
Carbon Bee’s Smart Striker X brand agnostic system uses sensors to spot spray both green-on-green and green-on-brown weeds. The system can also blanket dose application everywhere and then increase the dosage when a weed is detected, or modulate dosage based on biomass, weeds, or crop conditions. Weeds, crop and disease are detected up to 3 meters in front of the boom, generating a real-time map.
Ecorobotix’s ARA Field Sprayer. The unit features AI-equipped cameras beneath a 20-foot box-like implement. A computer vision system directs 56 nozzles to spray weeds individually — as small as 2.5 inches. The box shields the spray from wind.
“There are many crop-specific algorithms that tell the nozzles, ‘Just hit the weeds,’ or ‘Just hit the crop,’” says Chriss Wozniak, Ecorobotix’s U.S. marketing communications manager, noting operators can just select the crop, too. “You can even use nonselective herbicide and put a little buffer around each plant.”
The system’s AI learns with each pass, making it more effective at identifying weeds over time. For now, the ARA Field Sprayer works for onion, carrot and lettuce crops. There are beta algorithms out for broccoli, cauliflower, sugarbeets, sugar beans, canola, some corn and potatoes. Wozniak says they’re working on corn and soybeans.
Solinftec’s Solix. The robotic targeted applicator is fully autonomous and doesn’t look like any other ag sprayer.
“There’s a 40–foot boom on the front, with eight cameras across detecting where those weeds are for targeted application,” says Taylor Wetley, U.S. commercial manager at Solinftec. Compared to other targeted spray systems, the Solix moves through the field slowly. However, solar panels mounted on the top enable it to work indefinitely when paired with Solinftec’s auto refill station.
“We’re very used to the traditional, two- or three-pass program in corn and soybeans. We’re kind of flipping that on its head,” Wetley says. “The platform is acting as a rolling residual in the field, getting on acreage early and often to spray weeds when they’re small in order to reduce the seed bank over time.”
Variable-rate application systems
Raven’s Augmenta field analyzer. The system leverages a single multispectrum camera and computer vision to automate prescription application rates based on field conditions. The system doesn’t target the spray, but rather analyzes data and adjusts predefined rates accordingly.
“It looks out the full width of the boom to identify the target areas where we need to make adjustments,” says Kendal Quandahl, Case IH’s precision segment lead. “It increases or decreases the rate based on crop conditions.”
While spraying, the system documents crop conditions for later analysis. It can be used for in-season nitrogen, fungicide, plant growth regulators, herbicide (burndown) and harvest aids.
Agco systems. The company offers two variable-rate application systems: PTx Trimble’s Weedseeker 2, for green on brown, and Precision Planting’s Symphony Vision Rate, which adjusts rate based on crop conditions.
“Cameras ‘see’ the weeds, determine the severity — size and amount — and adjust the spray rate,” says Bryce Baker, marketing director at Precision Planting. Optional system configurations (either Spot or Rate) can either shut off application completely or adjust the rate to its lowest setting.
“Weedseeker 2 can provide significant herbicide savings by turning the nozzles off when no weeds are present in fallow ground,” Baker says, noting that it sprays anything green.
Agrifac’s AiCPlus. Its camera spraying system uses a complicated algorithm to gather, sort and label field data. The algorithm learns the pattern of when the weeds appear, and then uses cameras to sense individual weeds and control rate application accordingly. The system is best in paddocks with relatively sparse, but economically significant, weed populations.
DAT Ecopatch’s retrofittable sensor system uses deep-learning neural networks to predict and detect weeds. A computer analyzes images in real time and adjusts spray thresholds based on field conditions.
SprAI by DeepAgro. Designed for no-till production operations, the system can differentiate green from green. It can be installed on any sprayer to open and close nozzles based on field conditions. Independent modules consist of a camera, a computer and electro-valves for each nozzle.