AI in agriculture: new path to Africa's food security

AI in agriculture is being used to drive the sourcing of real-time data, perform predictive analysis, and run algorithms that optimize farming practices.
In South Africa, Kenya, and Zimbabwe, ITIKI, an innovative project is tapping indigenous environmental knowledge among African communities, integrating it with AI to predict droughts with better precision.
GSMA: AI in agriculture is poised to enable the deployment of innovative digital financial solutions such as credit and insurance products for millions of farmers.

In Africa, a continent of over 1.2 billion people, agriculture remains the primary economic activity, accounting for 17 percent of the GDP on average while offering jobs to nearly 60 percent of the population.

The bulk of the food produced in Africa or about 80 percent is, however, attributable to the effort of smallholder farms, where women provide much of the workforce. Unfortunately, these smallholder farmers continue to face multiple challenges, a scenario that perpetuates low productivity, reduced incomes, and the ever-present danger of hunger and food insecurity.

From Senegal to Cameroon to Uganda to Zimbabwe, smallholder farmers depend in on outdated farming techniques, strained access to capital, poor marketing systems, and near non-existent access to critical information that could improve agriculture.

Millions of farmers lack access to quality farm inputs largely due to their inability to get credit services, in particular, due to women’s inability to provide bankable collateral. Huge data gaps in their financial history make it even harder for potential financiers to determine their creditworthiness. Sadly, these woes hit women and youth in agriculture more than men.

In recent years, with the increasing negative impact of climate change, extreme temperatures, and floods, the future of Africa’s agriculture appears gloomy. Smallholder farmers are under increasing pressure from climate change, due to their dependence on rainfall and their ill-equipped coping mechanisms.

With relentless blows to the agriculture industry, the latest data paints a dire situation in the continent—50 percent of the population in Nigeria and 26 percent of the population in Kenya has insufficient food—a sad trend that leaves both nations facing serious levels of hunger. Many other countries are in worse if not similar state.

The role of digital and AI in Agriculture

Despite these challenges gripping agriculture in Africa, the rise of digital and AI applications in the farming industry is poised to turn the tide and secure the continent’s food security endeavors.

A new report by GSMA, titled AI for Africa: Use Cases Delivering Impact, digital technologies increasingly demonstrating the potential to unlock access to markets and services, offering a timely boost to Africa’s agricultural value chain while also mitigating the challenges and hurdles that smallholder farmers face today due to climate change.

The survey notes that AI in agriculture can drive the sourcing of “real-time data, perform predictive analysis, and run algorithms to optimize farming practices and improve crop yields.”

Additionally, digital and AI in agriculture innovations can strengthen means of market access for the players, and reduce production costs—a real pain point for millions of farmers across Africa.

And the revolution holds potential for investments in the years ahead, “For example, the agribot sector—comprising AI-driven robots performing agricultural tasks—is projected to reach around $337 million by 2030, marking a 21 percent compound annual growth rate from 2023 to 2030,” the GSMA report explains.

Already, AI in agriculture use cases are making waves in Kenya, Nigeria and South Africa among other leading economies in the continent, with the report identifying “digital advisory” as the most common usage area for AI.

Currently, the report adds, precision agriculture is bringing advisory services to the farm level by fusing “farm-specific data on farming with remote sensing data, climate and weather data, and domain-specific data.”

Governments, non-state actors, and agricultural agencies involved in addressing Africa’s food security have a chance to tap into the power of AI to help analyze vast amounts of data, identify patterns, and enable the creation of predictive models that better forecast food security outcomes.

AI in agriculture presents a new dawn where all actors can potentially step up the accuracy and timeliness of critical forecasts.

AI in agriculture use cases in Kenya

In Kenya, working with Microsoft’s AI for Good Lab in Kenya, healthcare services network Amref is developing an innovative spatio-temporal machine learning model that aims to detect malnutrition hotspots in Kenya. Once complete, this innovation will facilitate timely interventions to mitigate this challenge.

“The use of AI in this endeavor is not only about leveraging technology for predictions; it is about transforming data into actionable insights. Amref plans to integrate AI in analyzing comprehensive intervention reports spanning over five decades and real-time data from community health workers,” AI in Africa Meeting the Opportunity report by Microsoft states in part.

In the continent’s most advanced economy South Africa, ITIKI, an innovative project incubated at Central University of Technology, is championing a new platform that taps indigenous environmental knowledge among African communities and integrates it with AI to help predict droughts with better precision, offering alerts to farmers using mobile phones.

ITIKI’s approach makes use of indigenous knowledge by among others, reaching out to individuals to gather critical data on weather and climate. For instance, the project is tapping into communities’ knowledge of certain tree species’ blooming patterns, a phenomenon that usually indicates upcoming weather conditions. In African societies, such valuable knowledge is deeply rooted in local communities and is passed on from generation to generation.

Once these details are collected, ITIKI enhances this indigenous knowledge by fusing it with modern technology, tapping on a network of digital sensors collecting data on various environmental factors such as soil moisture levels, temperature, and similar others to make appropriate decisions that are critical to agriculture.

ITIKI’s innovative AI in agriculture solutions in use in South Africa’s Pietermaritzburg, and KwaZulu-Natal areas. In neighbouring Zimbabwe, smallholder farmers in Espugabera, Manica Province are making use of the technology while in Central Kenya, farmers in Mbeere, Embu County, are benefiting.

Additionally, “the growing availability of data and machine learning capabilities makes it possible to provide farmers with customized information based on field-level conditions. Examples of organisations using precision agriculture include ThirdEye in Kenya, Kitovu in Nigeria and Aerobotics in South Africa.”

In Kenya, ThirdEye uses flying sensors to monitor soil needs and identify pests/diseases in crops at an early stage, giving farmers timely and actionable details to improve their yields. For smallholder farmers in Nigeria, Kitovu combines spatial and satellite data to deliver precise agronomic advice, recommend optimal inputs, and ensure growers gain access to their crop health through updates on crop monitoring.

GSMA analysis shows that AI in agriculture is poised to enable the deployment of innovative digital financial solutions such as credit and insurance products for millions of farmers.

“By applying AI and ML to data sources such as satellite imagery of forms, climate data, and other farm-level information, organisations like Apollo Agriculture and mfarmPay in Kenya can use as collateral,” the report explains.

Used of AI in agricultural supply chain systems

Additionally, the use cases for AI in agriculture can extend to powering supply chain management, by among others, providing tools for real time monitoring and ensuring that players in the system maintain optimal storage conditions to cut post-harvest losses of farm produce.

AI innovation can also boost the value chain through applications that help streamline distribution and guarantee transparency on pricing and other relevant market conditions.

At the moment, startups such as ColdHubs and Koolboks in Nigeria are offering innovative products for farmers and market players, optimizing the storage conditions for perishable products such as fish and vegetables.

“Many organisations offer a bundle of services, for example digital advisory and DFS, typically aimed at digitizing value chain activities to provide holistic support to smallholder farmers.”

Overall, the usage of AI in agriculture tools and technologies depends on the availability and accessibility of a vast range of reliable data—agronomic, weather, and geospatial—to make the appropriate impact.

Across economies, hurdles such as the quality of data can often times limit potential AI-enabled services that are critical in optimizing the operation of Africa’s agriculture. “If research data predominantly reflects large-scale commercial farming practices, it will neglect the needs of smallholder farmers engaged in subsistence farming.”

What’s more, another unintended outcome might be the increasing targeting of mobile devices as the primary channel of reaching end uses. This method risks edging out non-mobile owners and illiterate farmers, thereby worsening productivity gaps.