The new national center of excellence ML4Earth (Machine Learning for Earth Observation) is developing novel AI methods for Earth Observation data and shaping a strong Data Science Community to tackle grand challenges to society from space.

Machine Learning for Earth Observation (ML4Earth) is a new national center of excellence lead by the Technical University of Munich (TUM). The center is resting on a collaboration of competitive partners. Together, the German Aerospace Center’s Remote Sensing Technology Institute, The Alfred Wegener Institute, and the Universities of Potsdam, Leipzig, and Bristol, are conducting research at the highest level on developing novel methods of Artificial Intelligence, applied to Earth Observation satellite data. By contributing to the European mission of building a Digital Twin of Earth, we tackle a grand challenge of our time: climate change and its consequences for society.
The ML4Earth project is also engaging in building the international AI4EO community to foster collaborative thinking and knowledge exchange.

Data Science Methods

Physics-aware machine learning, reasoning, uncertainty estimation, explainable AI, sparse labels and transferability, as well as deep learning for complex data structures.

Immediate application fields

European water storage, permafrost thawing, sea level budget, climate and earth system modeling (for example climate tipping points), soil parameter mapping, and multi-sensor segmentation (identifying farmlands, urban areas, and other classes in EO images).

Frequently Asked Questions

We are taking headshots at the moment. Very soon we will have the honor to highlight the motivated and capable team behind ML4Earth on this page.
The Data Science methods listed above mark the major fields of research. We will offer detailed insights, soon!
For sure! Our researchers will soon share stories to feature the scientific process, its success and failure, and perhaps even some personal insights.

Yes, you will find links to our papers here, soon. In the meantime, please see https://www.asg.ed.tum.de/sipeo/publications/ for a comprehensive publication list, also including our other research projects.

We are already looking forward to connecting with you and to offering you ready data products (benchmarks and more) and services (workshops, hackathons).

Get in touch

Lead: Prof. Dr. Xiaoxiang Zhu