A new weed-control system is upping the ante on broadleaf weeds and grasses that plague corn and soybean crops, protecting farmers’ potential profits in the process.
Called the Smart Spraying solution, the technology uses a sophisticated set of software and hardware to identify weeds, signal a set of nozzles to spray product and record the data in real time for farmers to review later at their convenience – all in milliseconds.
The system is the offspring of a joint venture between BASF and Bosch, according to Matt Leininger, managing director of North America Bosch BASF Smart Farming. The companies have been working together to develop the system since 2016.
“Bosch is known for their cutting-edge hardware, connectivity capabilities and work with OEMs (original equipment manufacturer), while BASF provides the weed-control know how and machine learning capabilities,” Leininger says. “The partnership is helping us deliver an exciting solution to weed control neither company could develop on its own.”
How It Works
In simple terms, the Smart Spraying solution works by photographing weeds — more than 1,000 per second – with cameras that cover the entire operating range of the sprayer. The cameras are positioned about every meter, Leininger says. Lighting units are also in place to eliminate any shadows.
Classic, deep-learning algorithms for computer vision are then used to recognize the different plants and distinguish between weeds and crops. Finally, the software automatically selects the type of herbicide to spray and signals the sprayer where to work.
“Afterwards, the farmer gets a map that shows what was sprayed, and even the weed density,” Leininger says. “The farmer can then go back to his retailer and talk about what’s happening out there in the field. We think this technology is going to open the door to a lot more detailed conversations.”
Vincent Orliange, project director for smart spraying at Bosch, says the system can be fine-tuned to fit individual farmer needs. The hardware specifications include up to
• 36 field camera units,
• 72 lighting units, and
• 8 section control units.
In addition, the system includes one master control unit, one connectivity control unit, two spray tanks for product and two sets of different-sized spray nozzles.
“The two tanks give you a lot of flexibility and the opportunity to do more than one job going through the field,” Leininger says. “As a for instance, if you want to apply a soybean fungicide and spray volunteer corn at the same time you can.”
Better Weed Control
The Smart Spraying solution offers farmers options for both pre-emergence (green-on-brown) and post-emergence (green-on-green) weed management. The system can be used in a variety of cultural practices, including no-till, strip-till and conventional tillage.
“You will be able to spot-apply a herbicide or make a broadcast application, using automation. Or, you can flip a switch and spray manually,” Leininger says.
Machine learning capabilities enhance the technology’s ability to help farmers achieve better product placement, application timing and product rates over time.
“Our goal is that herbicides are only used where they are really needed – the right herbicide, at the right place, in the right amount,” Leininger says. “By applying only where the product is needed, we reduce product use and the potential for off-target movement.”
In-Field Testing This Season
Orliange says the machine is in the proof-of-concept stage this year for U.S. growers, with 10 pre-series machines operating in fields at company and university research farms.
“Scalability is a big deal to us,” adds Leininger. “In real time, the technology must be able to handle being banged around going across fields and down highways. The system has to hit all the standards we have set for it at very high levels.”
Leininger says BASF and Bosch plan for a limited introduction of the Smart Spraying solution in 2023. A retail price for the technology will be available for consideration by this coming fall.
Leininger adds that customers will see a return on investment for the system within two to four years.