Three ex-SpaceX engineers are building a robot factory to make steel parts for AI data centres
Their startup, 1872, opened its first facility in Cincinnati this month. The goal: automate most of the steel fabrication process by 2027, starting with the frames that hold modular data centres together.
Key points
- Startup 1872 officially launched on 22 July 2026 with a factory opening in Cincinnati, Ohio.
- Three former SpaceX engineers founded the company, which uses AI software and robots to fabricate steel parts.
- The factory's first product is steel skids, rectangular steel frames that act as a moveable base for modular buildings.
- Primary customers are companies building AI data centres or small modular nuclear reactors.
- CEO Dan Summers says the target is roughly 80 percent automated operations by 2027, not necessarily 100 percent.
Three engineers who used to build rocket engines for SpaceX have turned their attention to something more earthbound: making steel parts with far fewer humans on the factory floor.
Their company is called 1872. It held a ribbon-cutting at Factory One in Cincinnati, Ohio, on 22 July. The name is a nod to the year steel became a mass-market material, which tells you how seriously these founders take the history of the stuff.
What does 1872 actually make?
Right now, the factory produces steel skids. A skid is a heavy rectangular steel frame, like a large sturdy pallet, that sits underneath a modular building and lets it be moved or repositioned. It is the unglamorous component every data centre needs but nobody talks about.
The company is targeting two types of customer: firms racing to build AI data centres (warehouse-sized buildings packed with computers that run artificial intelligence) and developers of small modular nuclear reactors (compact power plants designed to be built in a factory and shipped to a site). Both need precision steel work delivered quickly. That is the gap 1872 is trying to fill.
How much of the process is actually automated?
Most of it, eventually. The factory uses AI-driven software to plan and coordinate production, combined with industrial robots that handle welding and assembly. CEO Dan Summers, speaking to Ars Technica, is targeting roughly 80 percent autonomous operations by 2027.
He is not chasing 100 percent for its own sake. "We may achieve 80 percent autonomous operations, and we may decide that it makes sense to stop there because there's just diminishing returns to go to full 100 percent," Summers said. Practical, not dogmatic. Think of it like a dishwasher: you still scrape the plates, but the machine does the hard part.
Robotic welding at this scale is not hypothetical. Our story from 10 August on Tate Inc.'s shift to 58 welding robots found a 12-times jump in per-welder output and an end to costly rework, which is roughly the kind of gain 1872 is chasing.
Should you pay attention to this?
For construction or facilities managers sourcing components for a data centre project, a factory promising faster precision steel frames at scale is worth watching. For workers in fabrication: the company has not published staffing plans, so that picture is not yet clear, but automation in this space is accelerating fast. North American robot orders jumped 21 percent in value in Q2 2026, and the growth is no longer confined to car plants.
What matters most here is not the automation target itself but whether 1872 can sign the contracts to prove the model. The SpaceX pedigree is a good story; the order book will be the real one.
Common questions
Is 1872 a public company I can invest in?
No. 1872 is a privately held startup and has not announced any public funding rounds or stock listing as of its July 2026 launch.
Why focus on steel skids first?
Skids are high-demand, relatively standardised parts, which makes them a good fit for automation. Once the production process is dialled in on a simpler component, the factory can expand to more complex steel fabrication.
