Environmental Computational Science and Earth Observation Laboratory (ECEO)

3HD High Fidelity V2 4x
Devis Tuia Min

Prof. Devis Tuia

Environmental Computational Science and Earth Observation (ECEO)

Remote Sensing, Machine learning


Our mission

At ECEO we develop technology to monitor Earth with digital data. We use data from multiple sources (from the smartphone to the satellite) and develop algorithms enabling spatio temporal monitoring of our changing Earth. We study land processes, underwater ecosystems and ecosystems dynamics, for instance related to animal conservation and forest monitoring.

Research topics

1

Develop AI algorithms to make sense of digital data about Earth

2

Monitor endangered ecosystems like coral reefs or tropical forests

3

Map distirbution of species (animals and vegetal) and how they distribute in space and time, especially under climate change

Our key projects


1

DeepSDM

In collaboration with ETH, MILa and MIT (among others) we develop AI technology to map in space and time the distribution of (endangered) species


ETH MIT MILA Ohio State University

2

Coral reef monitoring in the Red Sea

We develop machine learning technology to map coral reefs with close sensing, support local monitoring programs and build capacity in the Red Sea Nations with dedicated workshops


Transnational Red Sea Center (EPFL) Univeristé de Djibouti Red Sea University

3

Treeline monitoring

Monitor alpine forests in Switzerland with 80 years of aerial photography. We develop AI tools to map and monitor the extend of the upper limit of Swiss forests and study changes in forest dynamics due to climate change


IDIAP institute Swisstopo UNIL /Jardin des plantes

Our results and highlights

1

Competion of coral monitoring mission in Djibouti (2023) and Erirea (2024), including training workshops for local scientists.
Release of our global map of Antarctica for blue ice

2

D. Tuia, elevation to IEEE Fellow

3

European Space Agency, Toward a Foundation Model for Multi-Sensor Earth Observation Data with Language Semantics, Open Space Innovation Program
Marie Curie European Doctoral Network WildDrone

Team & talents

Lab team size

18 people

Skills developed by the scientific team

Machine learning algorithms for understanding environmental processes
Low cost monitoring systems based on camera traps (Mammalps project with Swiss National Park), GoPro cameras (coral reefs monitoring) and fixed wings drones (WildDrone project, tested in Kenya and Namibia in 2025)

Regional and social impacts

1

We provide reproducible algorithms and scientific evidence at large scale of the evolution of several critical ecosystems of Earth

2

We collaborate with seveal services of the Canton (Géoinformation, Forets, Agriculture) to support them in their monitoring efforts, in particular to scale them up and automatize

3

Through collaborations with industry (Swisstopo, AXA), we develop technology critical to make their data analysis routiens more efficient and accurate

4

We train the next generation of environmental engineers in data proficiency, and prepare them to the future challenges of process monitoring, which are strongly influenced by big data problems and complex, multi-modal sterams

Perspectives and challenges

Main opportunities

Democratization of AI is a great opportunity to spread wider our technology
Great interest in Valais for AI technology helps us connecting with several key regional players

Future Partnerships

Expanding our coral technology in other reefs of the world, towards a citizen-led monitoring system
Extending our blue ice mapping technology to other problematics related to polar dynamics

Cookie policy
By continuing to browse this site, you agree to the use of cookies to improve your user experience and for the generation of visit statistics. You can customise the use of cookies using the buttons below.
My preferences