Artificial intelligence and farmer knowledge boost smallholder maize yields

La inteligencia artificial y los conocimientos de los agricultores ayudan a incrementar los rendimientos del maíz

Data-driven agriculture can boost smallholder production threatened by variable weather and climate change, but scientists need to work with farmers and governments. A decade of data collection and collaboration in Colombia shows what success looks like.

Farmers in Colombia’s northern maize-growing region of Córdoba had seen it all: too much rain one year, a searing drought another. In collaboration with a the government, a national growers association and researchers at the International Center for Tropical Agriculture (CIAT), they helped build and implement big-data tools that successfully increased yields in spite of the challenges presented by extreme weather and climate change.

The study, published in September in Global Food Security, shows how machine learning –  when applied to data from multiple sources, including, critically, farmers – can help make farming more efficient and productive even amid climate uncertainty. This can best be achieved when the scientists, producers and farmers organizations collaborate and work together, the researchers conclude.

 

“If one farmer provides data to a researcher it is almost impossible to gain many insights into how to improve management, On the other hand, if many farmers, each with distinct experiences, growing conditions, and management practices provide information, with the help of machine learning it is possible to deduce where and when specific management practices will work.”

James Cock

Co-author and Emeritus Scientist, International Center for Tropical Agriculture (CIAT)

Collected over a period of almost 10 years, Jimenez and colleagues analyzed the data and verified developed guidelines for increased production. Some farmers immediately followed the guidelines, while others waited until the recommendations were verified in field trials. Farmers that adopted the full suite of machine-generated guidelines saw their yields increase from an average of 3.5 tons per hectare to more than 6 tons per hectare – more than 40 percent. This is an excellent yield for rainfed maize in the region.

 

“Today we can collect massive amounts of data but you can’t just bulk it, process it in a machine and make a decision. The study shows in an honest way how we progressed, with successes and some failures. However, with institutions, experts and farmers all working together to reach a common goal, we learned how to surmount the difficulties. This spirit of working together was a key factor in reaching our goals.”

Daniel Jimenez

Data scientist and the study’s lead author, International Center for Tropical Agriculture (CIAT)

Not only did farmers obtained higher yields, but the guidelines substantially reduced fertilizer costs, and also provided advice on how to reduce risks related to variation in the weather patterns, including reducing the negative impacts of heavy rainfall.

Researchers from Colombia’s National Cereals and Legumes Federation (FENALCE) co-authored the study, which is part of a Colombian government program aimed at providing farmers with options to manage both weather variability and climate change.

Year by year, maize yields vary by as much as 39% due to variation in weather patterns. Small farmers in the past had to rely on their own knowledge of their crops and accept blanket recommendations often developed by researchers far removed from their own milieu. Now, by combining farmers’ knowledge and analysis of what happens on their farms with modern data sources of information on weather, soils and crop response to variables, farmers can better shield their crops against climate variability. They can also improve yields and reliably keep them higher.

In Córdoba, FENALCE, which compiles information on maize plantations, harvests, yields and costs, set up a web-based platform to collect and maintain data from individual farms. Local experts uploaded information on soils after visiting farms at various stages of the crop development, while IDEAM, Colombia’s weather agency, supplied weather information from six stations in the region. This allowed researchers to match daily weather station information with individual fields and the various stages of the growing season.

The researchers used machine learning algorithms and expert analysis to measure the impact of different weather, soil and farming practices on yields. For example, they noticed that improving soil drainage to reduce runoff will likely reduce yields where or when rainfall is lower, whereas doing the same in areas with a lot of rain will boost yields. This showed advice on crops needs to be site-specific. This contrasts with the blanket recommendations that were the only ones available before this the study.

The study highlighted management as the main cause of low yields. The research shows that by working with farmers and improving crop management, it is possible to increase maize production and food security and improve livelihoods, without large investments.

 

“The farmers felt part of the whole process and adopted the improved practices with confidence, After all, the guidelines were based on the information they themselves had provided.”

Andy Jarvis

Research Area Director and co-author, International Center for Tropical Agriculture (CIAT)

Human learning, too

Initially, CIAT and FENALCE designed a smartphone application for farmers to record soil and other data in the field but corn growers did not adopt the app. Although the web-based platform was used to compile the information, researchers and technical assistants had to visit the farms to help the farmers collect the data. This presents a problem for scaling up this type of exercise, and projects that follow this approach will need to address it.

Nevertheless, researchers are convinced that there are opportunities for increased data collection by smallholders, both by directly working with farmers and through technology.

“Much of the hardware and software for the future collection of cat may well come when the private sector becomes involved sustainable system for capturing, analysing and distributing information,” said Jimenez.

Future projects could incorporate apps already developed and tested for use by farmers. Furthermore, data collection by a whole array of technologies ranging from satellites, drones and low-cost sensors deployed in fields, and combine harvesters that accurately record grain yield at a micro-scale are all becoming realities in the developing world.

“In the future we can envisage every field being carefully characterized and monitored, turning the landscape into a whole series of experiments that provide data which machine learning can interpret to help famers manage their crops better,” said Cock.

Acknowledgements

This work was carried out under the ACLIMATE program (http://www.aclimatecolombia.org/http://odimpact.org/files/case-aclimate-colombia.pdf) with the financial support of the Colombian Ministry of Agriculture and Rural Development (MADR), the CGIAR Research Program on Climate Change, Agriculture and Food Security (CCAFS) under the project Towards a Digital Climate Smart Agriculture transformation in Latin America and the CGIAR Platform for Big Data in Agriculture under the community of practice Data-Driven Agronomy. Both CCAFS and the Platform for Big Data in Agriculture are carried out with support from CGIAR Trust Fund Donors and through bilateral funding agreements. For details please visit https://www.cgiar.org/funders/.