In April 2026, aboard Loft Orbital's Earth-observation satellite "YAM-9," Google DeepMind's vision-language model "Gemma 3" autonomously identified regions of interest within sensor data in response to natural-language queries. It is described as the first reported case of a satellite "finding what to look for" on its own, without a human analyst on the ground. The demonstration used a software package called "NAVI-Orbital," developed by NASA's Jet Propulsion Laboratory (JPL), with the model running on an Nvidia Jetson Orin AGX GPU carried by YAM-9. Details
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.