Early electrification of the manufacturing base is an interesting story. It isn’t just a story about replacing one sort of drive for another, but rather a process of iterative revelations that were only possible after each iteration had taken place. To keep the story short here I’ll summarize, but the details are almost as interesting as the high-level summary so I would recommend digging up books on the subject yourself if you’re the sort who likes to spend time with dead trees.
The Steam Engine
Industrialization really took off with the spread of steam technology. Steam power allowed enormous amounts of power to be harnessed directly from heat in the form of steam, and that steam could drive pistons or turbines. The details of which technology emerged when and which prevailed at different periods is not important, so we’ll skip it, but the main idea is that each large factory had its own steam power.
This meant that steam power was a fairly well distributed resource, at least at the granularity of factories. There was no central steam authority or city steam generator or anything. Each factory ran its own power operation, usually in the form of one huge steam engine per working line. These were massive, and were typically located at one end of the floor, the end of the building with the classic factory smoke stacks we’ve all seen in paintings and pictures.
A large shaft would run straight from the steam engine all the way down the line, with all sorts of belts and gears and drives attached directly to it. Downstream drive and clutch attachments that linked to further drives could be attached in the various ways needed to power whatever the manufacturing processes involved in the factory might be. Tooling a factory was quite a complex and expensive process, and changing anything was often a system-wide decision.
Because the power was shared, had a single source, and was all direct-drive, it wasn’t really possible to start and stop various machines independently. Clutching mechanisms were in use, but clutches wear out and are fickle, and the general case was that once the steam engine was engaged the whole line was engaged. Because of the all-or-nothing nature of the power, and the time required to build up steam in the first place (if you think boiling a kettle takes a while, imagine a steam engine), sudden stops due to mistakes or accidents in one place along the line were impossible or even unsafe, and horrific accidents were common compared to factories today.
The Advent of Electric Motors
Electric motors were viewed with skepticism by many. It didn’t seem particularly reasonable to introduce an additional layer of indirection to the distribution of power across a factory floor at first glance. Why use steam power to generate electricity only to power a massive electric motor to drive a power shaft other than it makes for a much better and easier to fine-tune clutching mechanism? Couldn’t this be solved with better mechanical clutches instead?
This argument is valid, and new technology is scary. As it turns out, even though early large electric motors did add some complication, they really were far better clutches and allowed for much better fine-tuning.
Let’s pause here and consider how hard the pitch for switching to electric motors must have felt to the customer and the marketer. Nobody at this time could envision any other way to deliver power to factory machines than single-drive shafts per line, because the only factory machines yet encountered were direct-drive mechanical devices. This made the marketing pitch incredibly difficult: only the very least interesting properties of electric motors, basically “better clutching and gearing”, could be used as levers to explain the proposition. Getting into any further discussion took a lot of time and energy and bored investors, bored the existing generation of designers, and so on.
Let’s blockquote that for emphasis, because it is central to the point I want to make:
Only the very least interesting properties of electric motors, basically “better clutching and gearing”, could be used as narrative points of interest to pitch the switch to electrification from steam power.
The Realization
Despite the difficulty of the pitch, fringe adopters were found, and some systems planning types realized something interesting:
- Thermal power scaled really well, which meant that generation of electricity scaled really well.
- Recruitment of that electric power by larger and larger electric engines was a pessimization.
That is to say, generation of electricity was such a win when it came to scaling that it actually made sense to separate electric power generation from the factory’s own operations. This meant that it made sense to go into business just generating electricity for many factories, which now opened opportunities for specialization and even competition within the narrow space of electric power generation. This allowed factories to simplify their own operations by outsourcing their power needs to providers that could allow everyone to benefit from economies of scale at a profit.
The second part of the realization was more of a revolution in industrial process, however. Once electrification had become normalized in the minds of the industrial leaders of the day — the ick factor over something being new had worn off — designers of factory machines started to realize that the single drive architecture of factory lines was really quite bad. Small electric motors weren’t just good, they were many times better than large motors, and running cables around the factory was a vastly superior process to arranging mechanical drives in sequence to pull power from a single shaft. Each work station could have its own power, its own little engines, and they could be switched on and off at will.
This led to new opportunities in flexible work schedules. It was no longer necessary for the entire line to be in operation, subject to wear and breakage, just to run one subset of machines. The factory could be re-tooled in increments while the rest of it remained in operation. Subsets of workers could come in and perform longer-running tasks independent of other sections, and so on.
The Outcome
The final impact of electrification was absolutely revolutionary in ways that we tend to overlook today, so far from the events. In retrospect we have a tendency to just imagine that there was steam power in factories, and they eventually got wise to electric power and did the obvious thing all at once. But no, the story is much more subtle than that.
Electrification was such a massive shift not only in process but in mindset that it could only happen in stages, and unevenly at that. Unless a person spends a considerable amount of time and mental energy with a set of ideas it is unlikely that they will come up with these extended impacts and see beyond the initial transition — and the sad fact is that only a tiny minority of people are capable of thinking through the extended vision of a truly revolutionary idea at all.
All of this is to say that when it comes to a radical transformation of deeply ingrained processes, the change must be introduced in stages and the pitch for the change must be condensed down to what the public is willing to hear, otherwise whatever is new and better will be rejected first on principle (“If it ain’t broke, don’t fix it” — typically a very sound impulse) and second on point of irritation (a handful of people enjoy listening to extended visions of a better future — most are viscerally repelled by this).
It is only after the weak argument has been entertained that a new technology can be adopted, and if it is sufficiently revolutionary in nature then this initial, minor adoption is enough to begin the incremental resolution of what would otherwise be an absolutely intractable chicken-and-egg problem.
There were plenty of people pitching the full vision at the outset:
- Small scale electric motors driving everything.
- Massive improvement in safety (enormous savings in liability and reduced loss of skilled workers).
- Completely new industry opens up for power generation.
- Incremental retooling becomes possible and other systemic improvements.
- Billions to be made in every direction! Clearly better!
But this message is heard by industrialists at the time as:
- We’re going to have to demolish your entire existing line operation.
- That’s a nice pinky swear about safety, but where’s the proof?
- Nobody is going to invest in a capital-intensive industry that has no reference point at all Also, do I have to wait until a power station exists before I can electrify or what?
- “Incremental retooling?” What does that even mean? Is this some jargon marketing buzzword? I hate buzzwords.
- Blah blah blah — everyone says that.
The lesson to take away is that revolutions in industry and process are very possible, but the revolution never leads. The revolutionary tail effects are only realized after the new technology has been adopted on its weakest merits first, and then processes themselves are re-imagined based on what is now possible.
Conclusion
Restating because this is the whole point of the story:
- A new technology must be adopted on its weakest merits.
- Revolutionary effects can only be realized later, after adoption, because only then can existing processes be re-imagined based on what is newly possible and the consensus mental model has updated.
Follow-on
The above is the background for a post I will publish later about a revolution we may be on the edge of, and the way I believe it will transpire based on what we can learn from historical examples of similar changes.