AI in Earth Observation – A Force for Good

The upcoming launch of the Φsat-2 mission is a prime example of the pioneering work ESA is doing in the field of AI in Earth observation.

But when it comes to AI, there is equal measure of excitement and fear. In this interview, ESA's Rochelle Schneider sets the record straight on how this innovative technology is improving access to critical information about the state and future of our planet.

The success of generative AI tools like ChatGPT has brought new questions about what lies ahead as the power and capabilities of AI continue to grow.

Fortunately, conferences such as AI For Good, which ESA frequently attends, demonstrate that rather than causing harm, AI can have a hugely positive impact on society and sustainable development.

As a machine learning expert, Rochelle Schneider, ESA Φ-lab AI Applications Lead, is well-placed to explain why AI can have a positive impact on Earth observation. She Rochelle has extensive experience in obtaining critical information from Earth observation data for disease prevention and child development.

Rochelle, how would you define the role of AI in society?

AI leverages computers and machines to perform repetitive tasks and functions for us that can be automated. This allows us to work more efficiently and focus on activities that require irreplaceable human roles, such as emotional intelligence, relationships, and intuition.

What do you think about the idea that AI will eliminate jobs?

There is no doubt that AI will transform society, and this will create immense opportunities for new careers. Therefore, we need to see AI not as something that will replace us in the job market, but as a tool that facilitates or enhances human capabilities and tasks.

We need to encourage the next generation of professionals to embrace the benefits of leveraging AI technology. For example, ESA's Earth Observations teaches young scientists and engineers to employ the tools of machine learning as part of their understanding of data and data processing.

Can you explain how AI will advance Earth observation?

Earth observation is greatly aided by AI. Some satellites are now equipped with AI computers for data filtering and processing, and there are many downstream applications that use AI to generate important insights for end users.

Computer vision amply demonstrates the contributions that AI can make. We extract many features such as vegetation and coal mines from satellite images. Computers have been doing this for a while, but traditionally a fixed algorithm or model was used to identify the desired object.

With AI, models continue to evolve and learn as new data arrives, making a huge difference in terms of the need for automatic detection. With vast amounts of Earth observation data generated every day, AI automation is an essential aid to Earth monitoring.

For example, ESA's Child Connectivity project with Giga assesses young people's access to the internet. As part of this, we used AI algorithms to recognize school buildings from Copernicus Sentinel 2 data, and in Brazil alone we were able to identify 65,000 schools that were misplaced in the UNICEF database.

The field of forecasting is also becoming increasingly important. In another major effort with UNICEF, our climate data AI model was able to predict dengue fever outbreaks a month in advance.

How will the rise of AI impact the commercial space sector?

Commercial Earth observation is growing rapidly, and AI is playing a key role here as well. A large number of companies have been established in Europe in recent years, many of which are supported by the ESA InCubed program.

These companies leverage AI to tailor satellite data for their customers and provide real-time information to government departments, disaster response agencies, insurance companies, and other industry stakeholders.

The soon to be launched Φsat-2 will be equipped with AI processing of images from multispectral cameras. Satellites already have built-in AI applications that remove clouds from images, create city maps, and detect ships, but ΦSAT-2 has the potential for private companies to upload applications and run them in space. There is also.

We expect Φsat-2 and other AI-enabled satellites to create many new business models and commercial opportunities for companies outside the space sector as well.

A final thought on AI: How do you think AI's contribution to Earth observation will develop in the future?

I think AI will be combined with many other emerging technologies in all sectors. For example, in Earth Observation, we are currently working on Destination Earth, an AI-powered digital twin of the Earth that monitors the impacts of natural and human activities and predicts extreme events.

Thinking about further steps, we can envision a digital twin of the Earth in the Metaverse, an immersive experience that provides enhanced interaction and simulation through augmented reality (a combination of augmented and virtual reality).

More generally, AI tools are increasingly entering the mainstream of ESA's operations and Earth observation activities, providing us with more insightful information for decision-making and critical management of our planet. will significantly boost our efforts towards a sustainable society.