DeepAgro has a solution called SprAI based on a combination of software, artificial intelligence, RGB cameras,
microcomputers and solenoid valves, which allows the detection weeds in different environments and applying the exact dose
of product only on the target.
It is a modular equipment, which allows it to be installed on any spraying machine of any brand and type. Each module has a
vision width of two meters, allowing it to adapt to any boom width.
A new product line is adapted to the machine, thus having a double working line, one for selective application and the other
for total application. In the new product line for selective application, nozzle holders are placed at 40 cm with selective
application tablets. The system is composed of a Ramsey valve to maintain constant pressure throughout the system regardless
of whether it opens 1 nozzle or the entire boom, guaranteeing the quality of the application. All materials for installation
are provided by DeepAgro.
Thanks to this solution, the producer saves up to 90% of the agrochemical to be applied and also allows to have an exact map
of how much and where the chemical was applied in order to have a quality certification of the application. Importantly, growers
are managing to reduce the environmental impact of each application by 90%, i.e. what used to have an impact of 100 on the
environment, the operator and the consumer, today has an impact of 10 (in relative EIQ values, which is the environmental
impact quotient used to evaluate agrochemical applications).
The following images show the detection capacity of small weeds:
In integrations with platforms such as unimap from acronex, the weed map of the lot can be visualized.
The worst thing about a lot is the escape of weeds, that is why we say that it is the evolution of green on green because the
equipment with sensors only detects plants with a reflectance (either NDVI or Chlorophyll) higher than that of the crop, thus
having escapes of small weeds. The great capacity of SprAI with its Weed in Crop function is the morphological detection with
camera and artificial intelligence, reaching the weeds at the right moment, with few growing points for better treatment,
thus achieving greater operability and efficiency of the equipment in the system.
Thanks to this morphological detection capability, we have been working very hard to differentiate grasses from broadleaves.
We have started with trials, but there is still time to get the product ready; it is a dream that we are getting closer to reality every day.
It is possible to differentiate green from green and detect weeds at any growth stage of the soybean and corn crops.
A key difference against competitors is that our system has been developed to work in no till production systems, where recognition conditions are tougher.
Allow high technology CPP (Crop Protection Product) at lower cost than low technology CPP in broadcasting application, reduce phytotoxicity on crops and improve yields.
Our system is competitively fast, providing high precision up to 18 km/h spraying speed, and it works day and night by means of LED lighting.