Welcome to Issue 24, where Google's Suncatcher satellite returns its first AI response
This was a quieter week. There was less news, and this issue has two short stories.
On the Space4AI side, Google reported on October 30 that its Project Suncatcher prototype satellite returned an AI response from an onboard data-center-class Google chip running its Gemma 4 model. We covered the satellite's launch in Issue 20.
On the AI4Space side, the UK Space Agency published a 73-page roadmap on October 27 for AI adoption across the UK space sector from 2026 to 2036. Space Professionals Partnership Ltd wrote it for the agency and the Department for Science, Innovation and Technology. The authors drew on more than 100 stakeholders through surveys, interviews and workshops.
Specifics below.
Did you know?
Bits & Orbits Insights
Public dashboards on in-space compute, space autonomy, and the SpaceX lock-up
calendar:
Space4AI
Google's Project Suncatcher prototype satellite returns first AI response from an onboard data-center-class TPU running Gemma 4
On October 30, 2026, Google reported that its Project Suncatcher prototype satellite had returned the first AI response from a data-center-class TPU in orbit. Travis Beals, Senior Director of Paradigms of Intelligence at Google, announced the result. We covered the launch in Issue 20.
After the October 1 launch, the team spent several weeks on a commissioning checklist. The checklist included powering up the host and TPU trays and running early machine learning tests. During a ground-station pass, the team uplinked a prompt asking the onboard Gemma 4 open model to write a haiku about being in space. The poem came back on the downlink as the satellite passed by Svalbard, Norway.
Google then tested prompts for space-themed limericks and prompts that return telemetry on the technical status of the payload systems. The next test series will measure how the TPUs perform under thermal stress and radiation exposure in orbit. Beals wrote that confirming the TPUs can run workloads reliably in space is the step toward the next milestone, twin interconnected satellites in 2027.
AI4Space
UK Space Agency publishes Artificial Intelligence Roadmap for Space, 2026-2036
On 27 October 2026, the UK Space Agency published the Artificial Intelligence Roadmap for Space (AIRS). The document sets out a ten-year framework for AI adoption across the UK space sector from 2026 to 2036. Space Professionals Partnership Ltd developed the 73-page roadmap for the UK Space Agency and the Department for Science, Innovation and Technology. The authors consulted more than 100 stakeholders through an online survey, one-on-one interviews, an online workshop and separate in-person upstream and downstream workshops.
The roadmap shortlists eight AI technologies from a wider framework of 12 AI functionalities. They include computer vision, simulation and digital twins, agentic AI for behaviour and control, human-AI interaction and edge computing. The roadmap maps them against seven investment themes: space domain awareness, resilient positioning, navigation and timing (PNT), satellite communications, Earth observation, autonomous operations, launch support and a national space data and AI infrastructure. This mapping produces eight recommendations, including a national AI hub to coordinate AI-led activities, phased over three periods: 2026 to 2030 for demonstration, 2031 to 2036 for scaling and certification, and 2037 onward.
The document says the UK is unlikely to compete on scale in hyperscale computing or constellation ownership. It identifies trusted, certifiable and resource-efficient AI for safety-critical systems as the UK's advantage. The immediate priorities include building AI requirements into UK Space Agency funding and procurement, and developing AI assurance and certification frameworks with the AI Security Institute, the Alan Turing Institute and BSI-linked standards work. A further priority covers edge AI research in neuromorphic computing, TinyML and radiation-tolerant AI hardware.
Till next time,
Meta-beat Column of this week
Read also about the AI Pipeline that sits at the core, producing this Newsletter, including its ups and downs of this week: