Cocoa farmer Mr. Jonathan Tettey didn’t expect a simple community workshop to transform how his family runs their 17-acre farm, but it did.
He has been farming in Akwadusuo, located in the Atiwa West District of Ghana’s Eastern Region, since 1998, relying on age-old wisdom passed down through generations.
For years, he could predict the rains by observing the skies, trees, and the movements of birds and insects.
A picture of Mr. Jonathan Tettey on his cocoa farm
But over the past year, something changed.
“The signs stopped working like before. The environment changed, and so did the rain,” Jonathan revealed.
The traditional indicators that had guided farmers for decades were no longer reliable and Jonathan believes rampant galamsey activities in the district, especially over the last year, are to blame.
The destruction of forests and contamination of water bodies have thrown the local ecosystem off balance, making it difficult to interpret nature’s signs.
In 2023, Jonathan took part in a training session organized by the Responsible AI Lab and the Artificial Intelligence For Sustainable Development (AI4SD) project.
The workshop wasn’t focused on technology just yet, instead, it aimed to revive indigenous weather forecasting methods, including cloud observation, tree behavior, moon phases, and animal patterns.
He didn’t keep this knowledge to himself.
Back at home, he trained his wife, Mrs. Abigail Teye, in the same techniques.
Mrs. Abigail Teye drying some cocoa beans
“My husband trained me
“Now, even if he’s not around, I can forecast the weather on my own. I can tell when it will rain or not,” said Abigail.
Together, they applied this knowledge to their daily farming decisions, from when to plant or spray to the right time to dry cocoa beans.
The results were immediate and impressive.
“After that training two years ago, my yields started going up. Now, I don’t get scared to farm or plant crops because I know when it will rain or not. We don’t fear to plant anymore,” Jonathan said.
The next phase came during a follow-up visit by the AI4SD team, this time with technology in hand funded by the French Embassy in Ghana with support from the Foreign, Commonwealth and Development Office (FCDO), the International Development Research Centre (IDRC), and the Artificial Intelligence for Development (AI4D).
A mobile application, developed in partnership with the Responsible AI Lab, was introduced.
The smart indigenous weather app
The app blends indigenous knowledge with machine learning, using farmer inputs and historical weather data to improve forecast accuracy.
Jonathan sees the app as a necessary tool, especially after the disruptions caused by illegal mining.
“I think this app will help. Due to galamsey activities late last year, it’s become a little harder to predict the weather. So having this as a backup could make a big difference,” he said.
The AI4SD AI in Agric theme also met with other farmers who had benefitted from the training for data collection as part of the app’s development.
The AI4SD AI in Agric Theme at Nyankumasi in the Cental Region for data collection
The team is scheduled to return in October to assess how the app is shaping farming practices in the district.
The AI4SD AI in Agric Theme at Mesrenyame in the Cental Region for data collection
