Satlyt conducted a successful test of its satellite data processing software after 13 months of development work with the University of Houston on a NASA-funded research project. The test, completed earlier this month, demonstrated how the startup's software can distribute computing tasks among orbiting computers based on communication availability and processing capacity.
The company's software is already deployed on three satellites in orbit. In the test, Satlyt created a simulated network with four computers, four users, and eight services, then assessed how to assign and run tasks across those machines while accounting for compute and communication constraints. One physical computer aboard a spacecraft ran the experiment, evaluating a simulated network of four computing hosts and returning results to Earth.
Founder Rama Afullo, who previously worked at Google and SpaceX, founded Satlyt in 2024 to turn spacecraft into virtual data centers. The goal is to have software that connects multiple satellites to share computing assignments, with users needing only "to define permissions, priorities, and operating limits," according to Afullo.
The test proved that Satlyt's software can generate task-assignment plans on hardware in orbit. However, the company identified a time-referencing issue in its software and found that one approach required more computing power than others. Afullo acknowledged that much work remains, including testing under various workloads and connectivity conditions, securing communication links, and handling connection failures across distributed spacecraft computers.
Satlyt was in the top five of TechCrunch Disrupt Battlefield this year. The company is also researching applications beyond low Earth orbit, including supporting 4K video streaming from the Moon through its NASA-funded work. While this test used NASA funding, the agency has not yet selected Satlyt for any particular mission or chosen it to deploy its technology.