Satellite view ©NASA

When humans and machines learn

Edgar Ssensalo, 27, is from Uganda and works as a Software Engineer. For him, artificial intelligence is the most important accomplishment there is – because machines that are able to learn can predict and shape the future, with sophisticated software and intelligent modelling.

: Friederike Bauer
: Visible Earth / NASA (2), Private, GIZ (4)

Edgar Ssensalo has always been fascinated by IT. ‘It got me through school,’ he laughs, because it was always the subject that got him top marks. He also focused on physics, chemistry and mathematics, but computers were always his passion. This enthusiasm led him to study Software Engineering at Makerere University in Kampala, one of the most renowned universities in Africa, where he obtained a Bachelor’s degree in 2020. He then began a career as a software engineer.

Edgar Ssensalo ©privat

To keep up with the latest developments in his field, Edgar Ssensalo is constantly upgrading his knowledge and skills. Two years ago, he attended a bootcamp on Machine Learning for Earth Observation (ML4EO). ‘I saw the course and was immediately intrigued. Fortunately, I was able to pass the prequalification assessment and got accepted in,’ he says.

Artificial intelligence for agriculture

Machine Learning for Earth Observation involves analysing and processing satellite data that not only tells us a lot about the current condition of the regions observed, but also allows forecasts about the future to be made through the use of modelling. This is particularly important for agriculture, which plays a significant role in Africa as it is the key economic factor. However, it requires specialists who are able to read and process the data. Just as ultrasound images are not straightforward to understand, special knowledge is also required to interpret satellite data.

This knowledge is taught at the bootcamp, a two-week intensive course supported by the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ) GmbH on behalf of the German Government. Participants learn how to process satellite data so that it can be used to draw intelligent conclusions. For example, whether a country is growing enough to be able to continue feeding its population in years to come, or identifying crop types using satellite imagery data, what condition the soil is in and what is the most promising thing to grow in a particular region. The data can also be used to record deforestation rates. ‘It’s fascinating, but at the same time very complex,’ says Edgar Ssensalo.

Satellite view

Sharing knowledge

What makes the bootcamp different is that a certificate of participation is not issued directly upon completion of the course, but only once an attendee has generated interest in the subject among five more people. ‘The knowledge needs to be shared.’ Edgar Ssensalo has trained three IT colleagues and two friends who were much less familiar with computers and software. This increased his own knowledge, too.

African farmers also benefit indirectly from this training because people like Ssensalo can use the models to develop apps for them. For example, apps to check which seed grows best on their land, or to retrieve weather data to find out the best time for harvesting. ‘There are a whole host of ways to use data like this effectively,’ says the IT specialist.

Bildergalerie Accordion

In the future, he would like to continue working in areas where artificial intelligence can be applied in agriculture. In 2021, for example, he took part in a competition held by the Radiant Earth Foundation. The non-governmental organisation was looking for ideas on how artificial intelligence could be used to benefit the Sustainable Development Goals (SDGs). Specifically, it needed to develop a model capable of identifying crop types in South Africa.

Fighting hunger

Edgar’s creative ideas know no bounds. At some point, he would like to advise governments or companies on artificial intelligence, for instance for long-term agricultural planning. With his support, they could then draw up more precise plans for the agriculture of tomorrow – and thus allow farmers to work more efficiently. After all, better yields are instrumental in fighting hunger and malnutrition. ‘The future lies in artificial intelligence,’ he says. ‘Not just here, but worldwide.’