NASA Enlists Public To Clean Up Space Telescope Data
NASA has launched a new citizen science initiative called Artifact InSPECtor that asks members of the public to help scientists distinguish real astronomical signals from data errors captured by two major space telescopes. The project invites volunteers of any age to review telescope imagery and flag anomalies that could otherwise be mistaken for genuine cosmic objects.
The effort centers on data from Euclid, a space telescope built by the European Space Agency with NASA contributions, and NASA’s upcoming Nancy Grace Roman Space Telescope, which is expected to begin science operations soon. Both observatories are designed to capture light from millions of distant galaxies, though they will observe different regions and densities of the sky. Together, the missions aim to help researchers investigate the expansion of the universe and the nature of dark energy, the poorly understood force believed to drive that expansion.
Each telescope relies on an instrument called a spectrograph, which breaks down light from galaxies into a spectrum of colors, similar to how a prism separates sunlight into a rainbow. These spectra allow scientists to estimate a galaxy’s distance, determine the types of stars it contains, and study supermassive black holes at galactic centers. However, raw telescope data often includes “artifacts” — false signals caused by factors such as light reflecting off telescope hardware, cosmic rays striking detectors, or electronic quirks in the camera system. NASA compared these artifacts to a smudge on a smartphone camera lens or glare from sunlight distorting a photograph.
To address this, astronomers have developed artificial intelligence tools intended to automatically identify and remove such artifacts. According to NASA, these AI systems still struggle with accuracy when working with data from newer instruments, which is where public volunteers come in. Participants will be trained to recognize artifacts in real Euclid data, and eventually Roman data once it becomes available in early 2027. Their assessments will be used to refine the instructions that guide the AI models, improving their ability to separate genuine astronomical signals from noise.
Citizen science programs have become an increasingly common tool in modern astrophysics, particularly as telescopes generate data volumes too large for scientists to review manually. Missions studying galaxy formation, exoplanets and cosmic phenomena have long relied on crowdsourced human judgment to complement machine learning systems, since human pattern recognition can still outperform automated tools in ambiguous or novel cases. This hybrid approach, often called “human-in-the-loop” machine learning, allows researchers to accelerate discovery while maintaining data quality standards.
The broader push to understand dark energy has intensified in recent years, as instruments like Euclid and Roman are expected to map cosmic structures with unprecedented precision. Because a small fraction of contaminated data could subtly bias large-scale cosmological measurements, ensuring clean datasets is considered essential to drawing accurate conclusions about the universe’s expansion history. Projects like Artifact InSPECtor reflect a growing reliance on public participation to support the vast data-processing needs of next-generation space observatories.
NASA noted that no scientific background is required to take part, and volunteers can contribute using a smartphone, tablet or computer. The agency said early participants, including children as young as nine, had already tested the platform.
According to the agency, the project also includes visual examples showing how AI models mark suspected invalid pixels in telescope images, which volunteers are asked to verify or correct. This story is based on a press release from NASA Science.
