1 minute read

“ AI is yielding optimal consumption and production levels with vertical green farms, eliminating waste, and vastly improving yields and resource efficiency.” — AI XPRIZE

My Take

Earth observation satellites (EOS) are often used for agricultural crop monitoring. How to get this working for developing nations without the resources of European countries who do this already is a challenge. You may want embedded image recognition programs into mobile phones so that farmers can track and get feedback on crops. This requires very efficient models which can be preloaded with data and run on small devices. Just as importantly, you need to gather preferences from local end users about what information they need.

AI can be used to track deforestation happening in real time in remote communities. Sometimes seeing the change in trees is easy but working out why it is happening requires complex pattern recognition. For example, forests could be cut down because they are a planted crop such as palm oil or lumber. Forests could be destroyed by weather events, illegally cut for lumber, fire wood or most commonly cleared to make way for agriculture. So AI and ML methods need to be used to figure out which of these causes is at play and what the impacts of each really are.

Each day up until the AI for Good Summit in Geneva on May 15 I’m writing up a thought on how Artificial Intelligence could impact on each UN Sustainable Development Goal. (Go to the first post.)

Update: So I’m a bit behind, but the summit is fascinating and I’ll be filling in my thoughts as I attend sessions and think about it in the coming weeks.

Mark Crowley has no official affiliation with IBM, XPrize, ITU or the UN. The views and opinions expressed here are entirely his own.