Artificial Intelligence (AI) can solve many of the problems confronting the African continent today, but researching and developing solutions in this regard appears to have been limited, at least until recently. This may now significantly change as Google formally opened its AI centre in Accra, Ghana.
At the new AI centre, the diverse team in Artificial Intelligence and Machine Learning, has a mix of personalities that includes experts of African descent, working on building AI-powered solutions to real-world problems, including helping communities in Africa and beyond to improve their lives.
The centre’s primary objective is to research into developing the general framework and theoretical basis for solutions that organizations may then implement in solving specific problems. From health to agriculture, Google’s AI tools can be used to solve real life problems, provided the right collaborations are made with those who can advance those solutions.
For instance using TensorFlow, a tool that has been open sourced by Google, a team of researchers from the University of Pennsylvania and the plant pathology team at IITA-Tanzania, developed an artificial intelligence model that can be deployed on mobile phones to monitor crop disease. It uses Artificial Intelligence (AI) and Machine Learning (ML) to diagnose, in real time, crop pests and diseases, and has been said to accurately diagnose leaves damaged by the two cassava viral diseases—Cassava Mosaic Diseases (CMD) and Cassava Brown Streak Diseases (CBSD) and by red and green mites.
“We open source code so that everybody can just go out there, take the code we publish and use it to build all sort of things,” said Moustapha Cisse, head of Google’s AI Research Lab in Accra, during an interactive session in Accra last week. “In fact, we have seen many times, surprising ways of using code that we open source.”
As Cisse explained, the AI centre tries to advance the foundational aspects of artificial intelligence, and use those advances in different fields. There are members of the AI team in Accra who are working on different impact areas such as health, and there are people interested in using AI to improve various aspects in agriculture. There are also people interested in using AI to analyze satellite imagery to support census and help policy makers in making more informed decisions.
This was emphasised by John Quinn, an AI researcher at the centre, who noted one of the many applications of artificial intelligence was using satellite imagery to determine population census, and also using it to aid emergency services in places with limited population data.
According to Google, its focus is how AI and ML can be used for social good. It says by working with partners from such diverse fields as medicine, transportation, environmental groups and small businesses, it can help to evolve AI and ML tools to meet real-world challenges. Currently, there are existing projects under the Google AI for Social Good program, and many of these can be replicated to solve problems on the African continent as well. These include:
- Flood prediction: Floods affect up to 250 million people, causing thousands of fatalities and inflicting billions of dollars of economic damage every year. Google has developed a system that combines physics-based modelling with AI to produce earlier and more precise flood warnings.
- Earthquake aftershocks: Google partnered with Harvard researchers to apply AI to seismic data, and created a model that — while far from fully accurate — can now do a much better job than previous models of predicting where aftershocks will occur. It says existing predictors are little better than chance.
- Healthcare and biology:
o Developed an algorithm to predict heart attacks and strokes simply from images of the retina — no needle or blood draw required!
o Google researchers have helped doctors detect the spread of breast cancer tumors — the doctors and machine learning system are better working together than either is alone.
- Environment, agriculture, and natural science:
o Researchers at Makerere University used TensorFlow to help farmers identify disease in the cassava plant, a major food source in the developing world.
o A dairy farm in Waynesboro, Georgia is using TensorFlow to keep cows healthier and more productive, similar to another project in the Netherlands.
Some of these existing examples cut across healthcare and agriculture, two areas of significance in Africa. With this new centre, researchers can now explore more options in domesticating these existing models to solve some of Africa’s nagging (development) problems. One may already have been done for detecting some Cassava diseases, but many more solutions are yet to be explored.
CALEB OJEWALE
