Artificial intelligence is a foundational technology that has the potential to help address humanity’s greatest challenges. We are now thinking about where to go next to maximize that potential. Today, we are announcing Project Suncatcher, a new research project that enables day-scale machine learning in space. We are working backwards from this potential future to explore how interconnected networks of solar-powered satellites equipped with Tensor Processing Unit (TPU) AI chips can make the most of the sun’s power.
Inspired by other Google moonshots like self-driving cars and quantum computing, we started doing the foundational work needed to one day make this future possible. We are excited that this is a growing area of exploration. Our initial research, shared today as a preprint paper, describes our approach to satellite constellation design, control, and communications, and initial learnings from radiation testing of Google TPUs.
Our next step is a learning mission in partnership with Planet to launch two prototype satellites by early 2027, test our hardware in orbit, and lay the foundation for a future era of large-scale computing in space.
