Science1 distinct publisher2 min readPublished
A permanent base at the lunar south pole will put landers, rovers and crews on the same terrain. Knowing where each of them is has to come first. The satellites meant to answer that start launching in early 2027.
The Scientist · Science desk

Compiled by The ScientistSomething wrong?How this is made
Five satellites is a thin constellation by terrestrial standards. GPS operates 31 and Galileo 28 [5][6], 59 between them [2], and a phone receiver simply picks whatever subset gives it usable geometry. The lunar service begins with one satellite and grows to five [11], roughly a twelfth the number now overhead on Earth [4], with most of its signal pointed at a single site near the south pole [11].
Against that, Heckler's suggestion that the lunar system could eventually beat Earth GPS rests on one specific error term: lunar radio waves do not cross an atmosphere, so they arrive undistorted [12][13]. The advantage is real, though narrow. The reporting carries no accuracy figure for the south pole service and no dates for satellites two through five, so how often a rover will see enough of them at once to fix its position is not something this material answers.
LuGRE is worth being precise about. The NASA and Italian Space Agency experiment acquired GPS signals from the lunar surface in 2025, well outside the coverage those signals were designed for [8]. Acquisition is the hard first step and that was a real result. It also happened on the side of the moon that faces Earth, which is the only side where it can happen: Parker's point is that line of sight is the whole dependency, and the far side never has it [9]. The objection scales badly outward, too. On the source's own figures, Mars sits roughly 583 times farther from Earth than the moon does [3].
The difference is traffic, not difficulty, which is why Apollo is not a precedent here. Twelve astronauts landed between 1969 and 1972 on navigation that looks crude beside GPS [7], and Heckler's account is that those missions never demanded fine spatial precision [c7b]. A permanently staffed site does, because the vehicles have to resolve each other and not just the ground [2][3]. The result is a stack of systems, not one standard. Earth GNSS handles the ranges it reaches, lunar satellites fill in where it does not, and onboard sensors are integrated across both [14].
The sequencing looks deliberate. The first navigation satellite is due about five years before the outpost is meant to be occupied [1], and boots are expected on the surface well before that [16]. Houston's stated ambition is unglamorous by comparison, a live map with moving dots on it [15], but a real-time picture of where everything is turns out to be the same product as the collision-avoidance layer.
Ranked by verification strength, evidence, and original report placement.
NASA plans a permanent outpost called Moon Base, beginning around 2032, to house astronauts for extended stays, complete with habitats, power systems, transportation and other infrastructure supporting what the agency calls an enduring human presence.
NASA is working toward a system of communications and navigation services that would let astronauts and autonomous vehicles determine their location, their destination and the time needed to get there; the new era of exploration calls not only for spacecraft, landers and rovers but a way to coordinate activity on and around the moon.
Gregory Heckler, NASA's deputy program manager for capability development, told Discover: "All of these things need to know where they are at a given time so they don't bump into each other."
The moon is some 240,000 miles from Earth, and at that distance signals from Earth navigation satellites become faint and difficult for receivers to detect.
GPS is a constellation of 31 satellites operated by the U.S. military, and according to NASA its radio signals let smartphones locate themselves almost instantly to within a few feet.
The European Union manages a separate constellation of 28 satellites, called Galileo, and several other countries have similar arrangements.
Distinct publishers with included, body-backed reporting in this cluster.
1 article · August 27, 2026
Follow any of these and your For You feed starts watching them — no settings page required.
product
Starling navigated without GPS using the cameras it already had1 distinct publisher
science
Earth's dragged spacetime is now measured to 0.1%, and quintessence models feel it1 distinct publisher
science
NASA's Starling fixed its own orbit by sighting other people's satellites1 distinct publisher
product
A five-watt nuclear heater on Blue Ghost turns lunar night into a procurement problem1 distinct publisher
Evidence-backed comparisons of source perspectives and observed adoption signals. Read the methodology
Which Builder, Operator, and Investor concerns the observed source mix emphasized—not a truth score.
Evidence, demonstrated adoption, hype gap, incentives, and confidence are assessed independently, each on its own current evidence. How these are measured.
One outlet, walking NASA's own paper trail
Every load count and distance in this story — 31 GPS satellites, 28 Galileo, 240,000 miles, five lunar relays — comes from Discover citing NASA and ESA reference pages, and the only two people who speak are NASA staff describing the program they run. That is a short, honest chain to the primary source and also the whole of it: no contractor, no outside navigation engineer, no one asked whether early 2027 is real. The physics claims travel well on that basis; the dates do not.
One receiver experiment, zero relays in orbit
Strip the schedule out and what has actually happened is a single event: LuGRE picking up GPS from the lunar surface in 2025. The constellation that the headline turns on exists as a contractor assignment and a launch window. Even on paper it is five satellites against the 59 that GPS and Galileo field between them, pointed mainly at one crater rim, so the first years of service are a pole, not a moon.
'Lunar GPS' is still a metaphor
The phrase does a lot of unpaid work. GPS means global, continuous and metre-accurate; what launches in early 2027 is one satellite serving one region, and the suggestion that it could beat Earth GPS is Heckler saying 'potentially in the long term' with an atmospheric-distortion argument and no accuracy figure behind it. Discover deserves credit for hedging in the same paragraph and for spending most of the piece on the awkward parts — far-side blackouts, sensor fusion, receivers straining at faint signals. The overshoot is in the framing, not the reporting.
The program describing the program
Both voices in this story hold jobs inside the capability being justified, and the reference material behind it is published by the agencies whose budgets depend on Artemis continuing. That does not make the engineering wrong — the line-of-sight problem is geometry — but it shapes which questions never get asked: cost, slip risk, whether five satellites is enough. Intuitive Machines, the party with revenue at stake, is named and never quoted, which at least keeps a vendor pitch out of the copy.
Solid on the constraints, soft on the calendar
We would defend the technical spine of this without much hesitation: signals thin out with distance, the far side has no line of sight to Earth, LuGRE happened. The dates are a different matter. Two of them — crewed landing inside two years, extended stays around 2032 — are stated as intentions with no independent check, and one is explicitly conditioned on things going smoothly. Read the sequencing as NASA's current plan, not as a forecast we can stand behind.