German Research Center for Artificial Intelligence is working on building an AI system that would utilize Copernicus Sentinel-2 data to goals of SDG’s. Patrick Helber from the Centre threw light on it at the ESA’s ɸ-week.
At the ESA’s ɸ-week, held at Centre of Earth observation in Frascati, Italy, between 12-16 November, the core theme was merging AI with satellite data for innovation and delivering benefits to a large number of people.
Population growth and rapid urbanization in developing countries would put a burden on the existing infrastructure in the already overcrowded cities. This will also lead to numerous problems including the dearth of resources and polluted water.
In order to deal with, the first step is a comprehensive mapping of urban habitats, predicting the explosion in population and analyzing it.
In scenarios like this, there is no other method other than AI to efficiently analyze data.
“Through our research, we can, for example, run over a Copernicus Sentinel-2 image with trained artificial intelligence and the type of land is automatically classified.
“This, in turn, allows changes in land-use, such as urban growth to be monitored automatically,” says Patrick Helber.
The project involves programming the computer on how to recognize land patterns. Supervising the system and train the AI on how satellite images should be segmented. Once the AI learns it will go on with segmentation and classification of new data, believes Helber.
Distinguishing between an industrial area and residential areas is crucial for land planning. And the AI should be able to do it efficiently. It depends on whether we have RGB information, multispectral information or optical information.
“Spatial resolution of the image is also important. With Sentinel-2 images, the advantage is that we can observe the change in Earth”, adds Helber.
While we stand at the cusp of a mega revolution spearheaded by emerging technologies like AI, AR, VR and robotics and await the 4h Industrial Revolution, we need to continuously up the ante towards achieving sustainable development goals as well. The utility of Machine Learning and Artificial Intelligence in Earth Observation has immense potential and can provide solutions to many of our current problems.
AI chip in smallsat
ESA is looking at new ways to explore Earth Observation and deliver benefits to the citizenry using Artificial Intelligence.
ESA satellites alone generate about 150 TB data every day and processing and utilizing this data is a big challenge.
Using Machine Learning algorithms and AI, the vast amount of data can not only is effectively analyzed but also helps in future data prediction.
Grappling with climate change and environmental degradation is a mounting threat before us and AI could assist in providing relevant information quickly.
Josef Aschbacher, ESA’s Director of Earth Observation Programmes, is also optimistic about the potential of AI in beterment of society.
“We believe that artificial intelligence is set to take us into a new era for Earth observation and this can only benefit society”, Aschbacher says.
With innovations in AI and its increasing use in Earth Observation, the day is not far when a small satellite would have an AI chip. This is one of Aschbacher’s personal dream and ESA is planning to launch an experimental small satellite with an AI chip next year.