News & Events

News

2025. 06. 04 -

FTMC Team Shines in International Hackathon: Satellites and AI to Build a Healthier Future

Stefan Schneider, Jury member and special guest at the 9th Cassini Hackathon, Space downstream Entrepreneurship Officer at the EUSPA with the "AIr Aware" team: Dr Mehri Davtalab, Abdullah Khan, Sonata Pleskytė, and Dr Daria Pashneva. Photo: FTMC

From 16 to 18 May, the remote CASSINI Hackathon took place, where participants from 10 countries developed and pitched innovative ideas. The aim of the competition was to explore how the EU’s Earth observation programme Copernicus could be used to support healthcare.

In the Lithuanian category, the FTMC Department of Environmental Research team, named “AIr Aware”, won second place. That’s right - AIr, not Air - referencing artificial intelligence (AI), which plays a key role in the project.

The team earned four hours of professional consultation, sponsored by investors from BSV Ventures. The jury was mainly composed of business representatives, and teams had just three minutes to “sell” their idea in the most engaging way.

This task was handled by Dr Mehri Davtalab, alongside with teammates PhD student Sonata Pleskytė, PhD student Abdullah Khan, and Dr Daria Pashneva. Additional support came from FTMC scientists Dr Vadimas Dudoitis (Department of Environmental Research), Dr Wanessa Melo (Department of Functional Materials and Electronics) and the head from the Department of Innovation Dr Karolis Stašys, who took a role as mentors in Cassini Hackathon.

(Photo: Pexels.com)

Free and accessible crucial information

What is the project about? The researchers proposed the idea of an interactive mortality prediction map based on satellite data and machine learning.

Why does this matter? “AIr Aware” points out that air pollution is a major cause of premature mortality worldwide (12% of total global deaths and over 7 million premature deaths annually worldwide), with vulnerable populations facing the greatest risks.

“Around 55% of the world’s population currently lives in urban areas - a figure expected to rise to 68% by 2050. Cities are hotspots for air pollution. Moreover, according to the World Health Organization, 90% of the population is exposed to poor air quality.

Traditional monitoring stations are used to measure air pollution, but these are expensive and not available in every city. Our solution is to use freely accessible Copernicus satellite data, integrated with machine learning, to estimate mortality risk even in places without air quality monitoring infrastructure,” explains A. Khan.

Currently, there are only 17 automatic air quality monitoring stations in Lithuania.

(Abdullah Khan. Photo from personal archive)

By integrating atmospheric data (e.g., PM_2.5, NO₂, CO) with mortality statistics, demographic data, and socioeconomic indicators, the model will:

• Identify spatial and temporal patterns linking pollution to death rates;

• Highlight inequities in exposure and outcomes across different regions and populations;

• Predict future mortality risk based on current and forecasted air quality conditions.

Importantly, the system is expected to provide governments and public health agencies with an early-warning tool to target interventions, allocate healthcare resources, and develop environmental policies that protect the most affected communities.

(Sonata Pleskytė. Photo from personal archive)

A user-friendly website is coming

Although the hackathon focused on Copernicus, NASA satellites also provide pollution data globally - and it’s all freely accessible.

Team member S. Pleskytė says their model’s key strength lies in its precision and accessibility: any user can easily check the pollution and health risk in a specific location:

“We discussed a lot of data. This model, assisted by machine learning, will be very specific to a certain location because we also include social and economic factors. It could be Lithuania, Poland, or another country - it doesn’t matter.”

“The uniqueness of our model is that it also account for social and geographic inequalities by using two key sources of EU data to help identify high-risk communities. Reducing such mortality not only saves lives but also lessens the long-term strain on public health systems and supports societal well-being.

My colleagues have already spoken about Copernicus, and I want to add about Eurostat. Eurostat is the EU’s statistical office, providing harmonized and comprehensive demographic and health data, including mortality rates and socioeconomic indicators like population, GDP, and across EU countries. All this data enables the model to link air pollution-related mortality to actual health outcomes”, says D. Pashneva.

(Dr Daria Pashneva. Photo: Hernandez & Sorokina / FTMC)

S. Pleskytė adds that “AIr Aware” will also be useful to academia from various fields, such as medicine, chemistry, and physics. Not everyone has the tools to analyse satellite data - but the future website will make it easy and understandable for everyone. Click a location - and get insights on pollution, mortality, and more.

“By using machine learning, we can identify the relationship between all of these factors and air pollution-related mortality. Finally, we provide an interactive map using cloud-based GIS platforms. This allows users to zoom in, zoom out, click on a specific area, and access all the necessary information.

You can see where the risk is low and where the risk of mortality is high. This is very useful for policymakers, urban planners, and insurance companies, as they are also highly interested. Moreover, it’s very important for planning where to build hospitals, kindergartens, or schools,” adds M. Davtalab.

(Dr Mehri Davtalab. Photo from personal archive)

Experience in school air quality monitoring

So what’s next? “After winning second place, it’s really a validation that we’re on the right path. It gave us motivation to go deeper and explore further,” says A. Khan.

M. Davtalab notes that working with satellite data is not entirely new topic. She already has experience in GIS, remote sensing, and machine learning:

“We started using satellite data earlier in our department to help vulnerable groups like students. For instance, we modeled air quality - for Vilnius schools to assess the situation using satellite data and machine learning. We initiated this line of research in our department.”

Mentor V. Dudoitis - an expert in air quality, aerosol science and particle exposure - also believes in the project’s future growth. He shares insight from the hackathon:

“There was one team that wanted to install large air-cleaning filters in strategic urban locations. They thought that simply placing these big devices outdoors would clean enough air for everyone to benefit.

However, such projects also have their drawbacks. We need to find a viable solution - one that actually works and is not too expensive. Because behind every project, there should not only be value for the end user or client, but also some form of revenue.

I just tried to assist by sharing this perspective, and I said: ‘If you plan to continue with this project in the future, feel free to reach out and we can discuss it further.’”

(Dr Vadimas Dudoitis. Photo: Hernandez & Sorokina / FTMC)

The “AIr Aware” team thanks Dr Steigvilė Byčenkienė, Head of the FTMC Department of Environmental Research, for encouraging their participation. They also thank mentors V. Dudoitis, K. Stašys, and W. Melo for their valuable guidance.

„Karolis helped us to prepare for the presentation. He gave us some ideas and advices from a business perspective. He is very good with colorful examples and was also supporting our team a lot.

Wanessa really helped us before the pitch session. She was so inspiring! We had precise comments very valuable for our team. Especially in challenging moments right before the pitch,” smiles S. Pleskytė.

Written by Simonas Bendžius