Why didn't electricity immediately change manufacturing? | Hacker News Reader

Why didn't electricity immediately change manufacturing?

bbc.com

102 pointsjackgavigan22 comments

An article from someone who knows little about the history of electric motors. I recommend "Men and Volts - A History of the General Electric Company" on this.

Early motors had serious limitations. DC motors wouldn't maintain constant speed as the load varied, had major brush wear problems, and tended to require frequent brush adjustment. Sprague, and the unknown person who invented carbon brushes, fixed that. AC motors wouldn't start under load. Tesla, Scott, and Steinmetz fixed that. Insulating materials were not very good. They were either flammable and subject to aging (varnished cambric, wood, paper) or not oil-tolerant and subject to aging (natural rubber). This limited the power density to less than 10% of what ordinary modern motors achieve.

The materials problem was particularly difficult. We're used to having quite good materials available for almost all purposes - insulators which can handle high temperatures and aren't brittle, wire that isn't brittle, low-cost ball bearings that run for years without oiling, all with consistent, repeatable properties from batch to batch. That was not the case up to WWII at all.

They did have asbestos, though. General Electric Deltabeston Wire. [1] "Will not age or crack." This was a high-end product around 1920.

[1] http://www.ebay.com/itm/1936-GENERAL-ELECTRIC-DELTABESTON-MA...

Fantastic comment, would love to hear more

I could write more, but this isn't the place for an essay on electric motor history. Since many VCs are on here, I suggest "The Electrical Manufacturers, 1875-1900 A Study in Competition, Entrepreneurship, Technical Change, and Economic Growth", by Harold C. Passer. This is a Harvard Business School book. It's like reading about the dot-com era, a century earlier. All the little startups, the patent battles, the consolidation, and finally duopoly - Westinghouse and General Electric.

That book has Edison's business plan numbers for the first power station, Pearl St. in NYC. Overoptimistic payback calculations and all. The margins were quite high. Gas lighting was expensive and electric light was much cheaper to provide, so Edison could and did mark it up a lot. Electric lighting was very profitable, which sped deployment. Initial sales were slow, though. The bottleneck was marketing, not the technology.

[1] http://www.hup.harvard.edu/catalog.php?isbn=9780674423640

Much appreciated.

Great thanks

...wire that isn't brittle...

That's funny. I need to change regularly the audio connection from the computer to the stereo. It always ends making an annoying noise. I guess I could buy the parts and a soldering iron and making a decent one myself...

The sources for this article are provided:

Robert M. Solow - 'We'd better watch out' New York Times Book Review, 12 July 1987

Robert Gordon - The Rise and Fall of American Growth (Oxford: Princeton University Press) 2016 p546-7

Paul A. David - The Computer and the dynamo: An historical perspective American Economic Review, May 1990

Paul A. David and Mark Thomas - The economic future in historical perspective OUP/British Academy, February 2006

Erik Brynjolfsoon and Lorin M. Hitt - Beyond Computation: Information technology, organizational transformation and business performance Journal of Economic Perspectives, Fall 2000

All economists, no electrical engineers.

A lot of it had to with the efficiency of DC motors as well. Electrification of heavy industry never took place until the introduction of 3-phase AC motors which were able to provide the sufficient horsepower.

https://en.wikipedia.org/wiki/Induction_motor

No, steel mill drives in both AC and DC were used.[1] They each have their advantages. Speed control in the early days took additional rotating electrical machinery. Huge rotating machinery.

Diesel-electric locomotives mostly used DC motors in the 20th century. 3-phase AC synchronous motors with elaborate semiconductor control systems are now used in most new locomotives, but that's only in the last 20 years.[2] There are still thousands of DC-motor locomotives running on US rails.

High-speed elevators were DC for most of the 20th century.

Speed control of DC motors was, for a long time, easier than speed control for AC motors. The reverse is now true, because power semiconductors are so good.

[1] https://books.google.com/books?id=-thQAAAAYAAJ&pg=PA308 [2] http://www.republiclocomotive.com/ac-traction-vs-dc-traction...

TL;DR

>> Two economists, Eric Brynjolfsson and Lorin Hitt, published research showing that many companies had invested in computers for little or no reward while others had reaped big benefits.

>> What explained the difference was whether the companies had been willing to reorganise to take advantage of what computers had to offer.

"Communism is Soviet power plus the electrification of the whole country." - Vladimir Lenin

Therefore

Soviet Power = Communism - Electrification

And

Electrification = Communism - Soviet Power

> But given the huge investment this involved, they were often disappointed with the savings.

Well, makes sense. Pretty easy to guess, and not really worth reading the article for the one line answer to the title.

I think the article proposes that the reason is rather in the need to change your processes if you really want to take advantage of the new technology. Just replacing your main steam engine with electric one isn't going to bring real gains and you're giving away some control. On the other hand, it allows you to organize the factory anew - which takes time.

Something I only came across only in the last few years on how quite straightforward a steam line shaft shop would have operated. This guy is running one out of his shed: https://www.youtube.com/watch?v=9WXHNBMLZZM. If you watch this you can also see how straightforward it was to retrofit electric motors to the existing belt drive mechanism, but you can see why it was slow to adopt. If you take into the early adopter concerns, the mains reliability concerns and a general risk-averse outlook lightbulbs are a better sell then electric motors.

From the article:

> "By 2000 - about 50 years after the first computer program - productivity was picking up a bit."

I think what made computers and productivity with computers really take off was down to 3 things:

- The PC and continued drops in price and increases in power.

- Spreadsheets

- Internet

I would say your last point balances out the others to show why productivity did not increase as much as you would expect!

Nah, spreadsheets did that on their own. Have you ever actually found a usable spreadsheet?

Different levels of usability.

Spreadsheets as simple ledgers? Wonderful, assuming you know about using ranges and how to use $ (or not) as part of reference statements.

Even that level of complexity seems like wizardry to most office drones.

However a spreadsheet is weak compared to a true relational logic engine; a modern database. Databases /have/ had the decades of refinement and are very good tools. Yet for those tools you kind of nail it. Specialists are required to correctly harness that potential.

I think this article demonstrates the 80/20 principle: Just having a technology on the marketplace is 20% of the way there. The other 80% is solving small, but critical problems, needed prior to adoption.

I wish we could stop abusing Pareto's law …

Rule of thumb for the 80-20 principle: 80% of the usage of this principle in a discussion is bullshit and only 20% of people talking about it ever heard of Pareto ;)

I ran across this idea a couple of weeks ago in the book 'Machine, Platform, Crowd". I'm drawing out idea for presentations to my com Sci high school students.

Reply on news.ycombinator.com