Towards Programmable Manufacturing

Over the next decade, leading robotics manufacturers have publicly pledged to produce hundreds of millions of robots. Each of these robots will house dozens of precision-manufactured components. As things stand, almost all of them will be made in China.

While a Western shop quotes a precision part in days, China quotes in under an hour, at a fraction of the cost. In China parts are made faster, revised faster, and reordered faster. Chinese manufacturing moves at a rate the West has not matched in decades, particularly on the high-mix low-volume end.

As a consequence, Western robotics companies will source their components from China, and much of the venture capital now flooding into Western robotics will end up paying for it. The money raised to build the West's robots will be spent building the East's factories.

In our eyes, this is a dependence worth challenging. After all, physical abundance will belong to those who manufacture it.

To address it, we need to raise the rate at which high-mix, low-volume manufacturing in the West operates, from the moment a part is quoted to the moment it is delivered.

That rate comes down to two things: how fast a shop wins a job, and how fast it runs the job once won. The first is the time lost quoting and revising a design before a part is ever cut. The second is the time lost between jobs once a part is in production: a machine spends most of its life not cutting but waiting, being set up, programmed, fixtured, and inspected. This changeover cost is what makes Western shops slow on the floor. Both are levers that make this opportunity commercially valuable.

Machine shops are high operating leverage businesses, which makes asset utilisation, measured in operating machine hours, a crucial lever that translates directly into higher margins. To increase asset utilisation in high-mix, low-volume contexts, you have to reduce changeover cost. We see this as fundamentally a problem of technical orchestration.

This focus on reducing changeover costs and quoting time, coupled with the multiple arbitrage that results from technically transforming the operation, makes a tech-first roll up play distinct from its traditional PE counterpart.

Many are trying to solve this by selling software point solutions into machine shops. However, selling standalone tools from the outside has a ceiling, because both are at root coordination problems. The delay does not live in any single step but in the seams between them.

Our bet is that to solve these coordination problems it is necessary to own and govern the entire operation. To change how a machine shop runs, you have to run it.

This is why we founded Standard Frontier. We build autonomous, intelligent factories by gathering the West's most talented engineers to translate frontier AI and robotics into the industrial standard.

We acquire precision shops from owners seeking succession, and we transform them from within. We start aggressively where orchestration breaks down today: on the AI side in quoting, machine programming, DFM feedback, and scheduling; on the robotics side in machine loading/unloading, fixturing, and inspection. We own the operation and carry the risk. We buy to hold.

Some of the necessary tools already exist. Digital point solutions, sensors, and advanced machinery have all been developed independently. Much of the stack, though, will have to be built in house. Our first objective is to reach full productive autonomy in a single shop: one operation, run end to end by the system we build. Everything after that is replication.

In time, Standard Frontier will run an entire network of these shops as a single system. The goal is abundant production addressable in software: programmable manufacturing.