Everything was ready to go. Place, date and joint objective – all agreed. The 31 researchers in the AI4EO International Future Lab were to begin arriving in Munich from July 2020. Then the Covid pandemic threw a spanner in the works. Only four of the twelve researchers in the core team have so far made it to Munich. AI4EO kicked off anyway, largely in virtual form with much of the team working from home.

Since early July, researchers from the USA, Europe and China have been investigating AI techniques that can be used to analyse satellite data. Looking ahead, this could help contain natural disasters such as forest fires or even make cities more liveable. AI4EO stands for “Artificial Intelligence for Earth Observation: Reasoning, Uncertainties, Ethics and Beyond” and is one of three international future labs in which the Federal Ministry of Education and Research (BMBF) supports collaboration between international researchers on artificial intelligence.

Most systems go!

The AI4EO team quickly adapted to the new situation. “The team members who are already in Munich take turns to work in the lab and from home,” reports Dr. Diana Mancera Guevara, AI4EO Science Manager. They communicate with the other team members using online tools or video conferencing. “That has worked well so far and has already led to some initial work products,” says Mancera Guevara.

Since 16 November 2020, Diana Mancera Guevara has held the reigns in the AI4EO lab. She leads the four-person management team that supports the researchers, organises the project and takes care of the lab’s technical resources. One of the most important and perhaps most difficult tasks at the moment is creating a sense of cohesion within the team. “Some of my colleagues would like more face-to-face interaction for a stronger sense of connection and collaboration,” Mancera Guevara acknowledges. She regards this as something special and human that can never be replaced by a machine – or online video conferencing. “All the same, IT has helped a lot to stay in touch during these difficult times,” she notes.

Mancera Guevara expects the months ahead will be very demanding for the AI4EO team. “The biggest challenge is quickly adapting to the unpredictable changes that keep being made during the pandemic,” she says. “At the same time, we need to maintain communication and team spirit and ensure high quality in our work.”

Sticking to the plan

Out of the twelve researchers in the core team, just two are working at AI4EO in Munich so far. Professor Mrinalini (Nalini) Kochupillai is one of them. Her subject of interest is the ethical questions associated with using artificial intelligence in Earth observation. She tells us how she is increasingly immersed in the research at AI4EO. “I find it really exciting to work with the scientists on our team, exploring how the data we use is made up and what sets it apart.” In the long term, Professor Kochupillai would like to use the project to develop guidelines on the ethical and practical handling of satellite data “based on fairness, honesty and integrity.”

As well as on the scientific subject matter at AI4EO, however, Kochupillai also has an eye on the people themselves. It is her task as the team’s ethics officer to support the science management staff in strengthening team spirit and fostering greater group cohesion. “Feelings of isolation during the pandemic contribute to stress, which can negatively impact people’s work and creative output,” says Kochupillai. Together with the team, she is therefore exploring suitable – meaning science-based – options for workshops on team building, stress management and synergising. “We have already held one event of this kind during our first ever online Christmas party in December, and it was very well received,” she reports. More such events are planned.

In total, AI4EO has 31 researchers from the United States, Europe and China. The core research team consists of 13 guest scientists. These are supported by 70 ‘Beyond Fellows’ – students and postdocs who join the Munich lab for between three and six months under a fellowship. The team is headed by the AI4EO Director, Professor Dr. Xiaoxiang Zhu, and three professors from the partner institutions as well as the twelve members of the Advisory Board. The Board comprises professors from AI and Earth observation-related disciplines who provide advice and suggestions for improvement as needed and recommend young researchers for fellowships at AI4EO.

Interdisciplinary collaboration

A defining feature of the AI4EO team is that all team members work together across disciplines and specialisms. This is also what makes the project so exciting for many of those involved. “What makes it special for me is the diversity that our researchers bring to AI4EO with their different areas of expertise,” says Dr. Daniela Espinoza-Molina, Science Manager and predecessor to Dr. Diana Mancera Guevara.  “AI4EO brings together people from all over the world to work on current research issues and share their knowledge and experience. I find that inspiring and motivating.”

Diana Mancera Guevara shares the same enthusiasm for the research project tied to the hope that it can help make the world a better place: “I really enjoy working in science. And I hope that our society can use science and technology to the benefit of the people.” At AI4EO, she aims to work with her team members to create something that “has a positive impact on our world.”