LUMELITE formation: make several satellites work together

NUS’s LUMELITE-1, -2 and -3 project explores formation flying: several spacecraft maintaining a useful relationship instead of operating only as unrelated vehicles. The design describes inter-satellite communication, precise timing and relative ranging, with separations in the tens to hundreds of kilometres. The scientific or practical result would depend on the group’s geometry as well as the capabilities of each satellite.

The program is distinct from the already launched LUMELITE-4 maritime experiment and the LUMELITE-5 relay — open definition project. NUS’s May 2024 account still described a future formation launch, while a May 2026 paper by participating researchers discussed the formation satellites and fault-diagnosis work. This lesson treats the documented development and engineering goals as evidence, without turning an older launch forecast into a completed flight.

Absolute navigation estimates a spacecraft’s position in a larger reference frame. Relative navigation — open definition asks where another member is with respect to it. For a simplified one-dimensional example, one satellite is at coordinate 1,000 kilometres and another at 1,120 kilometres at the same instant. Their separation is 120 kilometres along that axis. A real three-dimensional calculation needs all components and a consistent reference frame.

Timing is essential to that comparison. If the first coordinate describes one moment and the second describes a later moment, subtracting them can mix actual motion with the intended separation. A coordinated system must therefore associate position information with timestamps and uncertainty. It also needs procedures for delayed communications or an unavailable member, rather than assuming every satellite always reports perfectly synchronized data.

Compare positions at the same time

A formation does not become reliable merely because each member can fly alone. Teams must consider how failures affect the group: a missing position report, unexpected drift, a power problem or a failed command may change what the other members can safely do. Simulations and ground tests let engineers examine these cases before committing hardware to a particular coordinated operation.

The 2026 research explores diagnosing faults from satellite telemetry — open definition. Such work can inform operations, but a diagnostic model’s performance on a test dataset is different from demonstrated autonomous behavior in orbit. The broader lesson is to evaluate both levels of a distributed system: can each spacecraft perform its own tasks, and can the group maintain the information and geometry required for their shared objective?