How did manufacturing get so accurate?

by 0shaqHnsy

I know today we have computers, but what first put us on the path to accurate, consistent machine manufacturing? Was it a direct consequence of the industrial revolution, a related event, or a separate happening altogether?

meshfillet

There isn't a singular event that can be cited in what modernized manufacturing, so much as a series of gradual improvements over the course of centuries, starting well before the industrial revolution and continuing directly through it. That's something that makes it an fascinating topic of study, since it tends to intersect with so many other technological history topics.

In antiquity, the expectation of anything manufactured was that it was either primarily a labor achievement(e.g. the pyramids, the Great Wall) or the work of a skilled artisan(pottery, metals, etc.). Machines existed, but they were simple, and a combination of limited access to resources and limited theoretical knowledge made it hard to develop more sophisticated machines. If you know how to properly train a technique, gaining proficiency in it is mostly a matter of getting the materials and putting in the hours. But the great artisans of antiquity lacked viable opportunities to train their ability to automate things and do manufacturing in the way we do it today. Everything was done bespoke, with some exceptions(e.g. reusable molds), and the training likewise tended towards making luxuries. With limited supplies of many materials, there wasn't enough of anything to waste, and there wasn't a great deal of academic theory that could be applied to your trade.

Regardless, over the course of lifetimes gradual changes took place: available goods would change according to knowledge and economic conditions. In general, the world has grown more knowledgable with time, though cases of burned libraries and lost secrets certainly exist. Technical advances often threaten power balances, making them a flashpoint for political struggle. But the application of knowledge is primarily limited not by a specific decree, but by access to resources and markets. Where trade routes opened up, more complex goods often followed, but if the trade disappeared a regression was likely. Changing factors like an excess of one resource and diminishing of another often lead towards finding a viable substitution, sometimes one that is more effective than the original formula - and so the cycles of the market itself have some impact on the growth of knowledge. It isn't as linear or hierarchical as a game's "technology tree" might suggest.

Many of the interesting specifics about manufacturing can be described in terms of developments in related fields, such as these examples:

  • Improvements in materials and metalworking techniques. Lathes are often cited as a critical part of this puzzle: Early lathes were for woodworking, and the progression to working metal came much later.

  • The development of smaller machines with fine metalwork - clocks, watches, jewelry and other relatively small objects. Having literal "nuts and bolts" available is a major prerequisite for a lot of machine concepts, and watches and jewelry motivate higher precision.

  • New energy and energy transmission ideas - steam power, electricity, etc. Many machine concepts involve relatively simple ideas with a lot of energy applied to them, like the machine press.

  • More engineering knowledge available in a consumable form. Basic theories from physics, even after their discovery, take time to proliferate, be used in experimental research, and their results summarized. All of these aspects of knowledge became much more accessible to the masses both before the industrial revolution(through Gutenberg's printing press) and after(with industrial technologies that supercharged the press).

When discussing a new automation process - and this remains true today - a tradeoff of cost, flexibility, quality, and speed comes into play. If you have no quality bar, you can get the wrong result instantly; if you need customization, you have to either use more humans or automate in a much more complex way. And a lot of automation developments are towards quality first, because guaranteed high quality is the thing that is hardest to achieve by throwing human labor at the problem, and therefore the most valuable thing you could leverage in a trade. Speed and cost are what we tend to think of first since fast moving assembly lines are visually impressive, and the ability to manufacture cheaply en masse is the most immediately noticable "revolution" of industrialization. But those kinds of developments tend to run in tandem with processes that improve precision and reduce defect rates. The products we build today are lighter and cheaper on average than similar examples from 50 years ago in large part because we have that additional precision available, and we can apply different materials to change the cost structure. And the promise of the industrial market remains that we'll find an even more ingenious allocation of resources, just around the corner.