AI-2: Mapping Terrain

Guiding Ideas

Theme:

Tell the students: they have been sent on an expedition, with a crew of researchers, engineers, and scientists, to fly over a remote forest and take photos of the terrain below. These photos will be used to code a dataset that will then be used by an AI, which uses unsupervised machine learning, to monitor and find different areas where illegal logging is occurring or areas of forest that have tree diseases. The AI uses unsupervised machine learning to get an idea of, with a large number of pictures, what the forests should normally look like. The AI then locates these troubled areas by analysing a dataset, made from the large number of pictures, looking for abnormal differences: coloration, the increase of sizes of desolate places, migrating animals… As the AI locates these different areas, conservation efforts can be directed there to help the delicate ecological systems in those places.

A partner called Terrafuse is using AI to keep track of areas that have had forest fires or areas that are showing signs of possible forest fires occurring. Which can greatly help lessen the impact of them. For more information on their activities, click the link below:

https://www.microsoft.com/en-us/ai/ai-for-earth-OOICloud