Luna-25 Took the Fast Route

Luna-25 arrived in lunar orbit on August 16, 2023. Chandrayaan-3 had been there since August 5. Both were aimed at the south pole. Both wanted to be first. The coverage, to the extent there was any, framed it as a race between a legacy space power and an upstart — which was true, and pointed at the wrong thing.

Here is the detail that stayed with me, and the clocks are worth stating carefully because the contrast usually gets reported wrong. Luna-25 launched on August 11 and reached lunar orbit five days later. Chandrayaan-3 launched on July 14 and reached lunar orbit twenty-two days later, having spent most of that time spiraling outward from Earth in a sequence of orbit-raising burns, each one borrowing a little more energy from a gravity well it could not afford to fight head-on. Measured launch to touchdown instead of launch to orbit, the numbers are forty days against about eleven. Either way you clock it, one of these took roughly four times as long as the other.

That wasn't elegance. It was a constraint. India's LVM3 could not throw Chandrayaan-3 straight at the Moon, so ISRO planned a trajectory where orbital mechanics did the work the rocket couldn't. Weeks of patience substituting for thrust. Roscosmos had a Soyuz-2.1b and no such problem: point, burn, arrive.

On August 19, Luna-25 fired its engine for a pre-landing orbit correction. The burn was meant to last 84 seconds. It ran 127. The spacecraft dropped into an orbit that intersected the surface, and Roscosmos announced the following day that it had "ceased to exist" — a formulation I have admired ever since for how much work it does in three words.

On August 23, Vikram set down near 69 degrees south. India became the fourth nation to soft-land on the Moon and the first to do it anywhere near the pole. The Pragyan rover drove around for ten days, ran a laser spectrometer across the regolith, confirmed sulfur in the soil, and went to sleep at sunset. It never woke up. That was always the likeliest ending — nobody survives a fourteen-day night at 200 below on a seventy-five-million-dollar budget, and nobody was pretending otherwise.

Seventy-five million dollars. Less than most movies about going to space.

The easy story here is Russian decline, and it is a real story: Luna-25 was the first Russian lunar lander since the Soviet Union flew Luna-24 in 1976, in development under one name or another since the late nineties, delayed so many times that ESA had time to attach a navigation camera to it and then withdraw the camera after the invasion of Ukraine. But that read is boring, because it's about one country having a bad decade.

The tempting story is the one I started writing, and I want to mark exactly where it breaks. Two trajectories, one impatient and one patient, and the patient one lands — alignment over force, measured in delta-v. It's a lovely reading. It does not survive Chandrayaan-2, which flew the same unhurried spiral in 2019, from the same agency, and crashed into the Moon anyway. Slowness saved nobody. And Luna-25 died of a fault in a one-time pre-landing burn; no amount of extra spiraling around Earth surfaces a bug in a maneuver that hasn't happened yet. The trajectory is an illustration, not the mechanism.

What actually separated the two missions happened between attempts. ISRO flew a failure, took it apart, and re-flew four years later with the failure engineered out: more propellant, strengthened legs, a target landing box roughly forty times larger, and rewritten guidance software designed around its own off-nominal cases. ISRO called it failure-based design, which is an unusually honest name for a thing. Vikram succeeded because it was built by people who had already watched a Vikram fail. Luna-25 was a program with no living flight experience, carrying twenty-five years of untested assumptions into a single burn with no second draft.

The variable is iteration cadence. An organization that flies often and fails cheaply accumulates corrections. An organization that flies once a generation gets exactly one chance to be right the first time — and being right the first time is not something institutions are good at. Compost failures into learning, with a launch manifest attached.

Three years on, the south pole is getting crowded. China returned far-side samples. American commercial landers have made it down and then tipped over onto their sides, twice, which sounds like a punchline and is closer to the point: the company that tipped over twice has flown more lunar landers in two years than Roscosmos has flown in fifty. India approved a sample-return follow-on. Luna-26 and Luna-27 continue to slide rightward on the calendar in the same patient way Luna-25 slid for twenty-five years.

I would like to end on something hopeful about scientific cooperation. Instead: the most useful thing produced by that August was a demonstration that you don't need the biggest rocket, you need the shortest gap between attempts — and the organizations least likely to absorb that lesson are precisely the ones that fly too rarely to learn anything from it.