ESA Gaia Mission Data Analysis has been nominated for the 2026 World AI Awards, recognising the role advanced data processing, algorithms and machine-learning techniques are playing in extracting scientific discoveries from one of astronomy’s largest observational datasets.
Gaia spent more than a decade repeatedly scanning the sky. By the time science observations ended on 15 January 2025, the European Space Agency’s spacecraft had made more than three trillion observations of around two billion stars and other objects in the Milky Way and beyond.
The spacecraft has stopped observing. The data story, though, is far from finished.
Behind Gaia sits an enormous processing operation designed to turn raw measurements into scientifically useful information about stellar positions, movements, brightness, composition and other characteristics. Increasingly, researchers are also applying artificial intelligence and machine learning to that information to uncover structures that conventional searches can struggle to identify.
Turning trillions of observations into a map of the Milky Way
Gaia launched in December 2013 with an ambitious objective: build an extraordinarily precise multidimensional map of our galaxy.
Rather than simply photographing the sky, Gaia repeatedly measured objects over time. Those observations allow astronomers to calculate how stars move, estimate their distances and investigate their physical properties.
Processing that information is the responsibility of the Gaia Data Processing and Analysis Consortium (DPAC), a collaboration involving scientists and software developers from more than 20 countries.
DPAC develops the algorithms, software and computing infrastructure needed to transform spacecraft telemetry into Gaia catalogues. Its work covers astrometric, photometric and spectroscopic processing alongside specialist analysis of objects ranging from multiple-star systems to minor planets.
The scale is unusual even by modern astronomy standards. ESA records more than 2.6 trillion astrometric CCD measurements and over 530 billion photometric CCD measurements during Gaia’s science operations.
Finding useful patterns inside a dataset of that size is precisely where sophisticated computational techniques become important.
AI helps astronomers find stellar families hidden in the data
Machine learning has already found practical applications in Gaia science.
ESA reported that scientists have applied artificial intelligence, algorithms and machine-learning approaches to Gaia observations to identify new members and subgroups within star clusters.
That sounds abstract until the underlying problem is considered.
Stars that formed together can gradually spread across space. Some move in similar ways; others do not fit the neat patterns astronomers might traditionally use to identify a stellar family. Sorting genuine cluster members from unrelated stars becomes difficult when researchers are working across enormous catalogues.
Machine-learning techniques can analyse combinations of position, motion and other characteristics to uncover relationships that may be difficult to isolate through simpler selection methods.
ESA says Gaia data has enabled scientists to identify previously unknown open clusters and reveal complex structures inside known ones. AI-based approaches have also helped researchers investigate unusual stellar families whose members do not exhibit the coherent movement normally expected from stars sharing a common origin.
Here, AI is not replacing astronomical interpretation. It is helping scientists navigate a search space that has become extraordinarily large.
Gaia Data Release 3 expanded the scientific search space
The richness of the dataset became particularly visible with Gaia Data Release 3, published in June 2022.
DR3 introduced expanded radial-velocity measurements, stellar spectra, astrophysical parameters and a much larger collection of variable stars covering 24 variability types.
That gives researchers far more than a catalogue of coordinates.
Scientists can combine measurements of position and movement with information about brightness, stellar chemistry and spectra. Computational models can then search those multidimensional datasets for classifications, anomalies and relationships that would be impractical to inspect object by object.
Gaia’s archive is also designed for external research. Scientists can query it programmatically, including through Python tools such as Astroquery, making the mission’s catalogue available for independent computational analysis.
The spacecraft is finished. The processing is not.
Gaia stopped collecting new science observations in January 2025, but years of data processing remain.
ESA currently expects Gaia Data Release 4 in December 2026, based on 66 months of observations. Data Release 5, which is intended to incorporate the complete mission dataset, is not expected before the end of 2030.
That timeline underlines something unusual about Gaia: some of its scientific legacy will emerge years after the spacecraft itself stopped scanning the sky.
As catalogues become richer, researchers will have a larger foundation for machine learning, statistical analysis and other computational methods. The opportunity is not simply to process more stars. It is to discover relationships buried across dimensions of motion, distance, brightness, spectra and time.
Graham Cooke, President of the World AI Awards, said:
“ESA’s Gaia Mission Data Analysis demonstrates how artificial intelligence and advanced computational techniques can help scientists extract knowledge from datasets that have reached extraordinary scale.
“Gaia has created an exceptionally rich record of our galaxy, and machine-learning approaches give researchers additional ways to identify stellar groups, detect patterns and investigate structures hidden within billions of astronomical sources. We congratulate the ESA Gaia Mission Data Analysis team on its 2026 World AI Awards nomination and look forward to the discoveries still to emerge from this remarkable dataset.”
The World AI Awards recognises organisations, individuals, products and technologies contributing to the development and application of artificial intelligence across industries.
ESA Gaia Mission Data Analysis represents a different side of that landscape. There is no consumer chatbot or conventional enterprise AI product here. Instead, algorithms and machine learning are being used as scientific instruments — tools for navigating an astronomical dataset too large and multidimensional to understand through manual analysis alone.
Gaia may no longer be observing the Milky Way, but its data remains very much alive.
Learn more about the European Space Agency and Gaia at esa.int.
Discover the World AI Awards 2026, explore the nominees and learn more about the awards at worldawards.ai.
