SciTech Pulse
Science

NASA Calls on the Public to Help Clean Up Space Telescope Data

NASA has launched Artifact InSPECtor, a citizen-science project that asks volunteers to help train AI to spot errors in data from telescopes studying the universe's expansion and dark energy.

Step by step

  1. 1

    Telescope captures galaxy spectra

  2. 2

    Data contains AI-confusing artifacts

  3. 3

    Volunteers label real artifacts by eye

  4. 4

    Labels retrain the AI tool

NASA has launched a citizen-science project, Artifact InSPECtor, that invites members of the public to help train artificial intelligence to spot errors in data from space telescopes studying the universe's expansion and , the mysterious force behind that expansion.

The project focuses on data from the Euclid space telescope, built by the European Space Agency (ESA) with contributions from NASA, and will soon include NASA's Nancy Grace Roman Space Telescope after it begins science operations in early 2027. Both telescopes use an instrument called a , which splits light from distant galaxies into a rainbow of colors called spectra, letting scientists determine how far away a galaxy is, what kinds of stars it contains, and details about the supermassive black holes at its center.

Before that data can be used, scientists must remove "artifacts" — signals in the data that come from something other than real astronomical objects, caused by light glinting off the telescope's housing, cosmic rays striking the detector, or quirks in the camera or electronics. Astronomers have built AI tools to recognize these artifacts, but the AI does not always distinguish them accurately in data from newer instruments.

Volunteers who join Artifact InSPECtor look at real telescope data from Euclid and learn to recognize artifacts in it; their work is then used to improve the instructions guiding the AI tool. Nine-year-old Maeve F., who tried the project, said, "It's really cool that we can help teach computers new skills."

Terms explained

The story so far

  1. Study of Nearly 3,000 Supernovas Suggests Dark Energy May Be Changing Over Time
  2. Scientists Buried 2,000 Pairs of Underwear Across Switzerland to Measure Soil Health
  3. ETH Zurich Feeds 100 Petabytes of NASA Data Into a Supercomputer to Speed Up Disaster Forecasting
  4. Wasps Are Disappearing This Summer, and Scientists Are Worried
  5. NASA Demonstrates Satellites Can Navigate in Space Without GPS
  6. NASA, IBM Release Open-Source AI Model to Study the Moon
  7. NASA Calls on the Public to Help Clean Up Space Telescope Data
#citizen science#NASA#Euclid#Nancy Grace Roman Space Telescope#dark energy
Rate this story

Related stories