The Brazilian startup Agrorobótica uses NASA laser technology in order to provide farmers with quick, clean and affordable soil analysis. Their tool evaluates up to 1,500 samples per day without chemicals, using Rover Curiosity´s robots that NASA previously used to search for water on Mars.
The so-called AGLIBS equipment is able to supply data on the amount of organic soil carbon, texture (sand, silt and clay contents), macro and micronutrients and pH. If other methods can also measure those indexes, what is the point?
Faster, cheaper and better
The big deal is that AGLIBS does it faster, cheaper and better. Fábio Angelis, CEO and founder of Agrorobótica, explains: “We use Laser Induced Breakdown Spectroscopy (LIBS) to warm up samples until 12,000 kelvins, as hot as the sun. Because of this, all material goes into the 4th state of matter, plasma, a type of cloud. So our AI software analyses those nutrients lights in quantitative and qualitative parameters.”
While conventional laboratories need about 15 chemicals for analysing, which can take 10 days in São Paulo or even 40 in the Brazilian Northeast, the AGLIBS system delivers more precise agronomic recommendations to farmers in just 2 days.
“Such information is important for farmers, because it indicates the correct dosage of agricultural inputs to be applied for soil correction, preventing their insufficiencies or excess as well as unnecessary expenses. Not to mention the contribution for the preservation of the deposits of correction minerals and non-renewable fertilisers,” adds Mauro Angelis, one of the partners of the company.
Normally, growers take one geo-referenced sample of 300g per 10 ha in order to identify concentrations of organic matter, macronutrients (phosphorus, calcium, magnesium and potassium), micronutrients (boron, copper, iron, manganese and zinc), pH and soil texture.
“Thus, the software builds a complete map of a farm. We add qualitative carbon parameters, which is new and very important for efficiency and reducing costs. Our technology is more sensitive and allows greater accuracy for better decisions,” say Fábio and Mauro Angelis.
High costs
The development of AGLIBS started in 2015, parallel to the establishment of Agrorobótica, as the group of researchers from the startup and Embrapa Instrumentation, part of the Brazilian public agricultural research corporation Embrapa, observed that many farmers do not analyse their soils due to high costs.
The LIBS method, an acronym for optic emission spectroscopy with laser induced plasma, is remarkable for the advanced onboard technology in comparison with other techniques.
According to PhD in physics and one of Agrorobótica’s partners Aida Bebeachibuli Magalhães, the technique allows the equipment’s operational capacity to total around 360,000 annual sample analyses, which are performed without generating chemical waste.
“Such volume is almost twice the number of analyses performed by large laboratories in Brazil that use traditional methodologies involving 13 processes,” the scientist asserts, stating that the new technology resorts to a single process.
With over 15 years’ experience in the use of optics and photonics, the Embrapa researcher Débora Milori explains that LIBS was deployed because it is a technique that analyses a sample with little or no preparation; it is a fast, low-cost system; and it can identify elements simultaneously in a small sample like 0.5 gram, for instance.
Milori explains that the LIBS system casts a high energy laser into the sample, generating plasma that emits light deriving from the atoms and ions that are present in the soil.
“The light emission from each element is like a fingerprint that makes it possible to identify the atom that is in the plasma. That is how we can quantify soil carbon, nutrients and contaminants”, the scientist clarifies.
Agrorobótica aims to go forward with its solutions. The next steps are crossing analytic data from soil and leafs and bring its AGLIBS device into the fields. Fábio argues the main challenge to drive the equipment directly to farms is Brazilian soil features.
“Brazilian soil is highly compacted. In other words, it is really hard to find a durable solution for taking samples procedures. We need 2 samples at least. One from 0 cm until 20 cm depth and others from 20 cm to 40 cm. That is a difficulty, however, we already deliver much value to our clients.”
Regardless, his team is already developing a large data base in order to improve AI software and to be capable of predicting relations between soil and leaf nutrition. “This is the next level for our researches. It can lead to much more efficiency and even healthier food,” says Fábio.