
AI sharpens sky surveys: Webb-speed data science now fuels Rubin Observatory insights
AI from UC Santa Cruz has dramatically sped up James Webb data analysis and is now being applied to the Vera C. Rubin Observatory. The Neo model, a conditional GAN trained on Subaru and Hubble images, removes atmospheric blur and recovers fine details, boosting galaxy morphology measurements by about 2–10× and turning Webb-scale processing into days rather than years. When used on Rubin data, the technology aims to sharpen ground-based images so they rival space-based quality, helping maximize the science return from Rubin’s Chilean sky survey without replacing astronomers. The work is GPU-accelerated (NVIDIA) and underscores AI’s potential to accelerate discoveries across major observatories.











