News

Fifteen days. That is the current final-assembly time for a Lijian-1 carrier rocket in the Nansha factory of Chinese Academy of Sciences Aerospace (CAS Space), according to a Securities Times report on September 27, 2026 — down from roughly three months at first flight. The number arrived wrapped in a television feature on rocket manufacturing, but for anyone who has ever watched an assembly hall compress a build cycle, it deserves to be pulled out and stared at. A 6x reduction in assembly time is not a productivity improvement. It is a redesign of what building a rocket means.

The report is careful about where the speed comes from, and the honesty is the most interesting part. The work on the factory floor is still recognizably artisanal: engineers performing installation, wiring, and precision mechanical joining by hand. What changed is everything upstream. Each stage segment of the Lijian-1 now builds to standardized, frozen configurations — every rocket draws the same drawings, fits the same fixtures, and follows documented work instructions per operation. Only the satellite adapter varies with customer payload. Standardization, not robots, bought the first order of magnitude.

With frozen configurations, the physical logic of the factory reorganizes. Where one crew once followed a single rocket from bottom to top over months, parallel workstations now advance multiple vehicles simultaneously, each station operating against fixed interfaces and known tolerances. The factory moves to a flow-line mental model borrowed from automotive and aerospace mass production: the vehicle passes to the work, the work does not gather around the vehicle. Annual assembly capacity at the Nansha facility reaches 30 rockets; a matching launch cadence has settled into roughly monthly flights. The bottleneck has migrated from the factory to the launch pad — the moment any production system achieves industrial adulthood.

Western programs ran the same movie decades ago in different theaters. When SpaceX compressed Falcon 9 production, it did so by freezing the product, vertical-integrating the supply chain, and converting the factory itself into an iteration target — build, fly, instrument, redesign, rebuild. CAS Space's Nansha numbers suggest a state-affiliated Chinese program has absorbed the same script, and the CCTV program of September 26 describing the commercial space sector's shift from bespoke development to batch industrial production confirms the doctrine is now official, not accidental. Satellites, in the same broadcast, were described as products of smart factories on moving assembly lines, their production logic colonized by consumer-electronics discipline.

The automation industry should read the fifteen-day figure with a specific ambition in mind. Rocket assembly has always resisted flow-line automation for legitimate engineering reasons: large lightweight structures, single-quantity batches, tolerances that compound over tens of meters, joinery processes qualified per-article. The Chinese answer so far — standardize the article, industrialize the process, keep human hands on the joint — is cheaper than automation and, evidently, fast. That is also, historically, the precursor condition under which automation becomes buyable: robots deployed onto a stable, repeatable process definition rather than invented around a shifting one. Expect assembly robotics, automated riveting cells, and torque-traceable tooling to be the next investment wave at these factories, because now there is finally a fixed process to hand them.

Launch cadence has an industrial echo beyond the sector itself. A monthly rocket from a 30-per-year factory changes what satellite operators can assume about rideshare availability, changes what constellation designers plan around, and changes how insurance and scheduling markets price access. Frequency is the product. The factory that assembles in fifteen days sells, in effect, certainty — and certainty at price is exactly what every earlier manufacturing revolution delivered when it reached a formerly hand-built good.

Three months to fifteen days is the kind of number that only sounds small to people who have never stood in a factory watching an assembly sequence fail for the ninth time. Whatever the next launch headline says, the quiet revolution was the paperwork: the day the drawing stopped changing was the day the factory started racing.

Buyers of industrial automation should resist the obvious misreading. The headline number does not mean rockets are now built by robots; it means a production system finally stopped changing its mind. Every earlier attempt to automate aerospace assembly ran into the same wall: a fixture designed for revision C is scrap when revision D lands, and requalifying an automated joint costs more than the handwork it replaced. Freezing the configuration first converts automation from a research project into a capital purchase, which is precisely the sequence automotive manufacturers followed decades ago and the one most aerospace programs still refuse to accept.

The cadence argument has a supply-chain echo. Thirty vehicles a year implies steady consumption of fasteners, seals, avionics connectors, and propellant fittings, and suppliers who can hold statistical process control across that volume earn a position that one-off craft production never rewarded. It also concentrates risk: a single supplier lot defect now propagates across an entire flight rate rather than being caught during a bespoke build. Expect incoming-inspection automation and traceability systems to follow the assembly line, because batch production without evidence is simply faster failure.

None of this is a claim about any specific program's reliability, which remains a function of test discipline rather than build speed. But the industrial lesson generalizes cleanly: the fastest way to make a hand-built product cheap is to stop redesigning it, and the second fastest is to instrument what remains.

Written by: Maxwell, a manufacturing-systems writer with over a decade in automation and production engineering, who has timed takt rates in plants across Asia and Europe and still believes the most underrated machine on a floor is a frozen revision level.