U.K. startup Shield Space is developing autonomous cubesats intended to protect satellites from interference or attack. The company says its longer-term goal is to deploy swarms of up to 100 small spacecraft, using lasers, microwaves or other defensive methods to protect Western satellites—even if contact with ground controllers is lost.
The plans follow concerns about close approaches by Russian military satellites to ICEYE-X36, a private radar satellite believed to provide battlefield intelligence to Ukraine. In May 2026, four Russian satellites reportedly came close to ICEYE-X36. One of them was reported in October to have passed within 500 meters of the same spacecraft.
Shield Space’s first planned demonstration, Project Broadsword, is scheduled for the fourth quarter of 2027 aboard SpaceX’s Transporter-21 rideshare mission. The test will use an 8U cubesat and a smaller 3U target craft. The company plans to maneuver them from 100 kilometers apart to within 100 meters, progressing through four levels of autonomy from human-supervised operations to approximately full autonomy.
The startup says it wants to build defensive capacity using relatively inexpensive, commercially available components rather than relying entirely on custom-designed systems. It has commissioned EnduroSat to build an autonomous 8U spacecraft with chemical propulsion and imaging capability, onto which Shield Space will install its software.
Shield Space co-founder and CEO Graeme Ritchie said the company hopes to demonstrate that 100 cubesats could perform a protection mission for less than £100 million. The spacecraft would be released from a larger carrier in orbit, with future plans progressing from two full-sized 16U cubesats to a carrier holding 10, and eventually a larger system of multiple carriers across different orbits.
Significant technical hurdles remain. Small spacecraft must accommodate propulsion, power, sensors, computing, communications and defensive systems within tight limits on size and weight. Shield Space also says it needs real mission data to train autonomous systems, because synthetic data alone is not sufficient for space operations. A military swarm would also require more complex decision-making than a civilian satellite-coordination mission, including compliance with rules of engagement.
Bleddyn Bowen, an associate professor of astropolitics at Durham University, compared the concept to the Cold War-era Brilliant Pebbles missile-defense proposal, which envisioned small autonomous interceptors. He said the idea could be feasible with sufficient investment, but would require years of development and testing, including ways to test kinetic weapons without creating space debris, as well as a buyer for the system.
Sources: Space.com.