This AI tool helps farmers in Africa deal with climate change

"Any disruption that you see in the growing conditions would be propagated into our models and then will show what would be the impact in the production," Racine Ly, director of data management at the pan-African research non-profit Akademiya 2063, told The Current's guest host Robyn Bresnahan.

"So basically you can have a map where you can know what are the most likely crops that you can grow in a certain area based on the growing condition trends."

The machine learning tool has thus far been very welcomed by African farmers

"If we are able to keep up to date with what is happening at least we will be able to protect most of our crops," Ghanaian farmer Deborah Osei-Mensah said.

AAgWa is operational in 47 African countries and has an accuracy rate of 94 percent.

"Sometimes we have countries where our predictions are diverging. In that case, what we do is we update the model and update the predictions based on that event," Ly said.

"There are two ways of increasing food production. You have your crop management. But another way of maybe increasing production is to reduce losses. That's where AAgWa is focusing on.”

The tool has thus far been praised by local NGOs.

"The work that Racine is doing is fantastic because on the African continent, especially in sub-Saharan Africa, it is very difficult to predict the yield that the farmers will have," said Canisius Kanangire, executive director of the African Agriculture Technology Foundation.

The executive also highlighted how important the fact that this project is local is.

"When it is done by Africans, we have two positive things that we see. First, it is adapted to the conditions of the African continent. Second, it is at least an assurance that those technologies will be sustained," Kanangire told CBC Radio.