Tanzanian farmers turn to AI to combat climate uncertainty

Tanzanian farmers turn to AI to combat climate uncertainty


In Tanzania, unpredictable weather is putting smallholder farmers at risk. Traditional methods of reading the land are no longer reliable, and crop losses are on the rise. Farmers are experimenting with AI-powered tools that provide hyperlocal weather forecasts and agricultural guidance.

Eric Mwandu, Rada360 agronomist, and William Karatibu conducting soil nutritional analysis to optimise agricultural productivity

Eric Mwandu, Rada360 agronomist, and William Karatibu conducting soil nutritional analysis to optimise agricultural productivity

William Karatibu is a lifelong bean and maize farmer who grew up reading ant patterns, clouds, and the wind to predict rainfall. But with heatwaves, floods, and delayed rains becoming more common, his traditional knowledge no longer works.

“I’ve been doing exactly what my parents taught me – following their methods just like they showed me. But recently, it became more difficult because the traditional ways of observation are less valuable with the current conditions. The methods my parents taught me are not working anymore,” says Karatibu.

Across Tanzania, climate change is hitting smallholder farmers hard. Average annual rainfall has declined over the past two decades, while extreme weather events are increasing.

Smallholder farmers face climate uncertainty – and a data gap

Around 65% of Tanzanians live in rural areas, and agriculture employs a significant portion of the workforce, contributing 25.3% to the country’s GDP. Most farmers operate on a smallholder basis and face challenges due to the lack of reliable data.

Essa Mohamedali, Tanzanian AI strategist and co-founder of the Tanzania AI Community, highlights the potential of AI to fill these gaps:

“One key insight is that AI in Africa is being driven by grassroots communities – passionate individuals, recent graduates, and first-time founders,” he says. “Tanzania has made great strides in connectivity, with internet coverage at 72% and more citizens coming online through mobile phones. Regarding data infrastructure, the data exists but is scattered and siloed.”

From crop failure to precision farming

Searching for solutions online, Karatibu discovered Rada360, a precision agriculture platform that uses satellite imagery and AI analytics to provide real-time crop monitoring and climate risk advice.

“I was experiencing so much loss that I tried to look for a solution. I went online and found out about Rada360 and understood that they were doing precision farming. To me ‘precision’ means high accuracy,” he says.

With the platform’s hyperlocal weather forecasts, Karatibu adjusted his farming practices and used the information to guide his decisions about planting and soil management. Neighboring farmers have asked him about his approach, and he says it has helped him make more informed choices.

Rada360 processes satellite and Earth observation data to provide insights that farmers can use for monitoring crops, planning for floods or droughts, and predicting yields.

The platform is one of four winners from the Adaptation & Resilience ClimAccelerator, a programme supporting local innovation for climate adaptation in Tanzania.

The platform is one of four winners from the Adaptation & Resilience ClimAccelerator, supporting local innovation for climate adaptation in Tanzania.

The climate cost of climate tech

While AI offers solutions, it also comes with environmental costs. Large AI systems consume significant energy and water, and building the necessary hardware can contribute to emissions. Mohamedali stresses the importance of embedding ethics into AI development:

“Ethics need to be baked into all AI training – technical and non-technical – along with more time to ponder and explore the philosophical side of the questions, time to ponder on what the outcome of an unethical solution looks like.”

Workshop participants working in teams on AI and climate assignments at Sahara Sparks, Dar es Salaam, September 2024.

Workshop participants working in teams on AI and climate assignments at Sahara Sparks, Dar es Salaam, September 2024.

AI literacy and local innovation

Programmes such as the Training Programme for AI-Driven Climate Change Solutions, launched by Climate KIC and Omdena, train Tanzanians to use AI to address climate challenges in their communities.

Mohamedali notes that building AI literacy fosters curiosity and innovation, while initiatives like Climate KIC’s Adaptation Innovation Cluster provide tools, training, and networks that help rural communities build resilience to climate impacts and explore local innovation.