Is Big Data a new input for farming?

Data have become new input in agriculture, enabling farmers to increase productivity levels. The structured information can, to some extent, improve results, like fertilizers, seeds, chemicals and water already do for growers.

But how? Two of the most relevant Big Data agricultural initiatives in Brazil can provide some answers. These initiatives are Embrapa Agricultura Digital (Digital Agriculture), in Campinas, São Paulo, and the “HUB de IA do Senai”, in Londrina, Paraná.

There are others too, mainly inside universities or commercialized by big companies. However, the two mentioned specifically aim to spread Big Data and further informatics tools all over the agricultural sector.

Collections of data

For example, Embrapa Digital Agriculture, a public institution linked to the Ministry of Agriculture, Livestock and Supply, which developed pioneer efforts in Brazil for both informatics and Big Data applied to farm activities.

The company has been collecting data about the sector in Brazil since 1985 and, currently, has got 240 structured data sets following FAIR (Findability, Accessibility, Interoperability, and Reusability) international standards.

To have a better idea; there are more than 100 terabytes of genomic information regarding several species, in just a single area. Meaning if you count data sets about climate, soil, market, society, production, etc, the sky is the limit.

“Big data can bring a strategic vision and generate knowledge to farmers, researchers, traders, and so on. Anyhow, we need to process and give access to this info to benefit farmers widely and within the whole sector”, says Stanley Oliveira, chief at Embrapa Digital Agriculture.

Digital agriculture for free

All this data is the beginning of several initiatives and ongoing solutions. The first example is Embrapa’s app in the Google Play Store store or Apple App Store which offers dozens of apps for free for many cultures and activities linked to all kinds of farms.

“The big producer generates a lot of data, the small one not so much. But Embrapa looks to all of them to make technology available. It is for everyone. The only requirement is an internet connection”, he adds.

Some of the favourite tools for Brazilian farmers are the apps “Plantio Certo” (or Right Planting, from the Portuguese) with more than 50,000 downloads, “Bioinsumos” (Bioinputs) with 10,000 or “Satveg” with 3,000.

Plantio Certo shows best planting dates

Number one, for example, shows the best planting dates for at least 60 crops by retrieving data from the Zoning Agricultural Climate Risk (Zarc), which is an instrument for risk management in agriculture.

The study is designed to quantify risks arising from adverse climatic conditions and allows each municipality to identify the best planting date of each culture, in different soil types and crop cycles, according to plant features.

This app, developed by Embrapa Digital Agriculture and its partners, is pivotal for data and information organizing that will support the Ministry of Agriculture for public policies to benefit farmers.

SmartSolos to identify soil types

Soil classification is now available for farmers all over Brazil. With the SmartSolos application, producers will be able to identify the type of soil of each plot in real-time, thus creating the appropriate destination for each area.

The technology presents the results as the user enters his data about the soil profile into the system. Firstly, the user creates a simple account and then makes a general description of their property by uploading soil and profile data and photos.

Information such as the date and geographic location are automatically updated by the system. According to Embrapa, an approximate classification is obtained at the end of the description given by the producer.

Embrapa´s API

Furthermore, Embrapa´s data lake has other functions as well. The public entity opens its API to other researching partners and business initiatives, such as companies and agtechs.

“Each year we promote an agtech program within our business model. It allows them to begin from a very rich platform. Probably the most complete in Brazil. After all, we have almost 40 years of official Brazilian agriculture data”, Stanley comments.

Due to this, Embrapa Digital Agriculture´s API has more than 100,000 requisitions per month. It means that partners and peers use their info 3,000 times a day.

“In addition to Big Data, we believe the most important new technologies for agriculture are AI (artificial intelligence), genetics, robotics, and Internet of Things (IoT). They are enabling tools that begin to show what digital transformation can offer to the world”, concludes Stanley.

Amazon genetic bank for bio inputs

Another interesting Embrapa project involves Big Data, bio-inputs, and the Amazon’s biodiversity. The initiative already has found at least 60 bioactives with technical and commercial potential to be applied in agriculture.

Embrapa Western Amazon and the Amazon State University collected micro-organisms from 7,000 kilometers of Amazon Rivers in order to check their potential to become valuable bio-inputs.

Using cutting-edge techniques of Big Data, genetics sequencing, and chemistry, scientists have analyzed these fungi and bacteria by their “genetic fingerprints” and compared them to a world genetic data bank.

After that, they could focus just on the microorganisms that bring agricultural results due to their genetic and biochemical features. This speeded up the process greatly and expanded the possibilities to make discoveries “at scale” of new bio-inputs.

HUB and anchor companies

The second institute, Senai Londrina’s AI HUB, began its operations in 2019, aiming for the main objective of promoting artificial intelligence technologies adoption included in agribusiness.

This means taking advantage of the massive amount of data on different aspects through analyses that make easier and better decision-making processes.

“The HUB´s main focus was to work with residency programs in AI. Our AI specialists could solve challenges identified by anchor companies using Big Data and artificial intelligence techniques”, explains Naira Pizzinatti, Innovation Consultant for agriculture at SENAI Institute in Paraná.

According to her, it has already solved challenges from more than 50 anchor companies and created more than 180 concept proofs during the last three years, based on 5 Terabytes from 30 data sources.

Classifying soybean

One example of their work is the tool for soybean classification through Big Data, Artificial Intelligence (AI), and Computer Vision. Demanded by a big cooperative in Paraná, Cocamar, Senai’s Artificial Intelligence HUB developed the system.

The apparatus captures grain images, extracts information, and trains the algorithm to monitor the acidity level and the concentration of chlorophyll. It is through this classification that the grains are evaluated to compose the purchase price that Cocamar will offer to the producer.

Step forward

Senai is planning to launch Habitat Agro Digital to integrate partners in the sector. The main goal is to connect the triple helix (ICTs, Government, and Companies) by exchanging data with each other to search for common solutions.

“Obviously, privacy and secret risks need to be respected and agreed between institutions, but Senai can be the partner to work on these alignments both from a technical and governance point of view in the integration of business data”, Kleber Canuto, Data Science Coordinator at Senai PR.

He mentions the General Personal Data Protection Law (LGPD, in Portuguese), which makes Brazil become part of the countries that have specific legislation for the protection of data and the privacy of their citizens

At the end of the day, all that structure could influence challenges for the entire sector, such as logistical optimization, quality inspection, models for defining the best moment for purchase and/or sale of agricultural products.