The Munich Technical University (TUM) and the German Aerospace Center (DLR) have launched a new initiative focussing on the development of artificial intelligence technologies in Earth observations. The Future Lab “Artificial Intelligence for Earth Observation: Reasoning, Uncertainties, Ethics and Beyond” (AI4EO) aims to use satellite data and Big Data analyses to create models of global urbanization, food supplies and disaster management. AI4EO was launched in May 2020.
The knowledge gained from AI-based Earth observation data can help before disasters through prevention and preparedness, as decision-makers are able to design sustainable cities accordingly. It can also help during and after disasters with response and recovery by, for instance, helping to bring forest fires under control. This is possible through satellite observation of certain geographic areas for a longer period of time, among others. Hence, AI technologies are not only expected to support achieving Sustainable Development Goal 11 “Making cities and human settlements inclusive, safe, resilient and sustainable”, but also address the full disaster management cycle. The amount of Earth observation data being collected is increasing rapidly and therefore AI technologies will become more and more relevant in the future to access such data efficiently.
DLR contributes to the Future Lab by providing decades of expertise in developing remote sensing methodologies and assessing data. The Future Lab AI4EO brings together a total of 27 researchers from 20 organizations, as well as 70 junior scientists, from 9 countries (United States of America, Europe and China) to foster the development of AI technologies in Earth observation. In order to reach this goal, research teams take a holistic and interdisciplinary approach, meaning that they focus on developing new Earth observation methods through AI technologies, and also investigate fundamental questions concerning AI technologies’ accuracy and reliability.
The initiative also addresses questions of ethics, such as whether and to what extent knowledge gained should be made public, as it could lead to improvements, however potentially also to abuse or misuse. Therefore, data protection and privacy will be addressed as well within this context.
The German Federal Ministry of Education and Research (BMBF) will fund AI4EO for three years with up to 5 million Euros.