Wednesday, May 18, 2011

Computer Numeric Re-Considerations (VII)

Amazingly, nearly a month has passed since the last installment in this series. It's time! My apologies for the less than regular posts in this thread.

In the last post on the topic of CNC and automation, I delved into, admittedly without great depth, the historic and social underpinnings of the development of automation in US manufacturing. At the close, I mentioned the great concern in US manufacturing since the days of Fredrick Taylor's Scientific Management: who's running the shop? Particularly in the arena of the machine tool industry, the person who operated the machinery controlled the rate of production in the shop, and 'soldiering', the deliberate pacing of production by the tool operators was a thorn in the side of management.

As industries grew larger and more 'rationally' managed, particularly following the upsurge in US manufacturing in world War II, and in particular, given the post-war development of new types of aircraft which required complex parts made to very high degrees of accuracy, Numerical-Control (NC)-cutting and forming machinery came to be seen as a necessity. The Air Force's own specifications for the products explicitly demanded their use.

Now there were two directions that this new technology might have taken. One was NC. Another was to use automated control systems as an augment to the skilled work of established craftspeople. This is the 'road not taken', which happens to also be the title of a chapter in David Noble's 1984 work Forces of Production.

An alternative system to NC was called Record-Playback, or R/P. To grasp the essential difference between NC and R/P, one can consider two examples: the Jacquard Loom, which used punch card technology to operate weaving looms; and the player piano, which seems superficially similar to the loom in terms of operating system, using a punched roll of paper instead of cards.

In the card-directed loom, we have essentially a digital, binary system, the punched holes in the card serving as go/no go controls for pins to pass through, or not. If the pin could pass through the card, it lined up with a wire to be lifted, with which a thread was lifted. To change the pattern, the card only need be changed out. The operator's task remained, however, unchanged, and, more to the point, the weaver's work, in the absence of any automated control, was seen only as a straightforward operation, easily replaced by the card. The weavers' skill set was, in short, devalued. The cards were produced by specialists, not weavers. Weavers acted to load and unload the machines, and change the cards.

The player piano, though using a punched card form of operating system, was designed around an entirely different philosophy of use. Those pianos make use of a device called a melograph, a device invented in the 18th century which recorded the fingering of a musician at a keyboard as marks on paper. Here's a somewhat amusing clipping describing the device from an 1887 edition of The New York Times:

Accusations have been leveled at machine based technologies, as debasers of human skill, for a long time, haven't they?

The melograph was adapted so as to employ pneumatic card punching of paper rolls, and these rolls were then incorporated into pianos. As the roll plays, the keys on the piano are activated in 'perfect' reproduction of the piece originally recorded by the melograph. Maybe not so perfect in practice though, as noted by Noble:

"But, in other ways, the melograph reproductions fell far short of human capabilities. Piano rolls generated notes rather than music, since they did not reproduce the subtleties of phrasing and dynamics. "At its best," one leading historian of player pianos observed, the music produced "was unequal to the efforts of an inebriated pianist on a much misused upright in the public bar." Thus, before long, player pianos were equipped with "manual expression controls," which enabled the operator of the roll-actuated piano to add his own dynamics and interpretations, while sparing him the need for digital dexterity."

The key difference between the player piano and automated loom technologies was that the player piano was based upon the real life skill of a musician. The player was not intended to eliminate that skill altogether, but to reproduce it as faithfully as possible, in order to multiply, magnify, or extend its range. It multiplies the skill of one musician and reduces the need for such skills on the part of other musicians.

R/P programming systems for machine tool operation were much the same as the player piano system - designed here to record the movements of a skilled operator fabricating a piece on a special machine, like a lathe or milling machine, designed to transmit the machinist's moves into a storage medium:

"The program was made, therefore, by "capturing" on tape the motions of a machine as it was put though its paces by a machinist, whose skill was thereby "captured" in the process. Here, as with conventional machining, the machinist interpreted the blueprint instructions and process sheets and manually made a first part (using a tracer stylus attachment to produce contours, if necessary). The program, therefore, was a record not only of the machine (and stylus) motions but also of the machinist's intelligence, skill, tacit knowledge, and judgment, which were embodied in those motions. Rather than viewing the possibility of human intervention cynically, as merely the chance for "human error," this approach viewed it positively, as the opportunity for human judgment, skill, and creativity. Reliant upon shop floor experience and cooperation, it was, by definition, limited to the capabilities of human machinists (and, thus, to more than 90% of metalworking applications)."

Once the copy was made of the skilled machinist's operation, that could then be used to actuate production machines, be they lathes or milling machines, etc., without the operator required to do much more than load, unload, and keep an eye on things. Just like a player piano.

Machinery was prototyped using R/P systems by large industrial concerns with specializations in electrical engineering, as these developments depended upon programmable magnetic tape systems, not punched paper. General Electric, in their laboratory in Schenectady, New York was one of the leading developers. Interestingly, there's a sort of a connection between GE and player pianos - a certain Kurt Vonnegut worked as a PR hack for GE in the late 1940's. Out of that experience, he wrote his first novel, with the title Player Piano in 1952. In a 1973 Playboy interview where he was asked about that work, Vonnegut stated:

"I was working for General Electric at the time, right after World War II, and I saw a milling machine for cutting the rotors on jet engines, gas turbines. This was a very expensive thing for a machinist to do, to cut what is essentially one of those Brancusi forms. So they had a computer-operated milling machine built to cut the blades, and I was fascinated by that. This was in 1949 and the guys who were working on it were foreseeing all sorts of machines being run by little boxes and punched cards. "Player Piano" was my response to the implications of having everything run by little boxes. The idea of doing that, you know, made sense, perfect sense. To have a little clicking box make all the decisions wasn't a vicious thing to do. But it was too bad for the human beings who got their dignity from their jobs." (emphasis mine)

Very insightful and prescient. Here's a brief excerpt from the novel:

"Strange business, this crusading spirit of the managers and engineers, the idea of designing and manufacturing and distributing being sort of a holy war; all that folklore was cooked up by public relations and advertising men hired by managers and engineers to make big business popular in the old days, which it certainly wasn't in the beginning. Now, the engineers and managers believe with all their hearts the glorious things their forbears hired people to say about them. Yesterday's snow job becomes today's sermon."

Hmm. I think that may be worth a read. More information on the rising influence and history of the managerial and engineering fields, BTW, can be found in Noble's earlier work America by Design, which I'm working my way through currently.

Make no mistake, the purpose of R/P technology was first and foremost the reduction in skill required for production. Noble again:

"The new technology...could serve as a "multiplier for the few outstanding machinists," thereby making possible the hiring of less skilled and hence cheaper machine operators - "an advantage," Orrin Livingston pointed out, "not to be underestimated."

R/P also offered other benefits:
  • errors on the machinist's part when recording could be erased and re-recorded, insuring an error-free master tape
  • recording can be stopped and started during the machining process so that dead time could be eliminated
  • recordings could be made a slow speed and then played back at high speed to take advantage of high-speed cutting tools
  • magnetic tapes did not suffer from dimensional change problems due to temperature, humidity, or mishandling near as much as other systems
  • use of the tapes reduced the tooling and set-up times required in subsequent production and translated into reduced inventory, since parts could be made cheaply on demand
Sounds great - so why didn't R/P technology develop further? There was a confluence of factors, but the most significant, according to Noble, was the fact that the operators still largely controlled production:

""With record-playback", Earl Troup pointed out, "the control of the machine remains with the machinist - control of feeds, speeds, number of cuts, output. Thus, management is dependent upon operators and can't optimize the use of their machines." John Dutcher agreed. He had already designed an automatic machine for grinding steel rolls, at the request of the Bethlehem Steel Company. "Bethlehem came to us," he later recalled, "complaining that operators were controlling production, determining the output - say, only eight finished rolls a day, no matter what." The steel company wanted GE to design an automated system that would give management control over output to increase it and at the same time eliminate worker "stints" (the worker-determined quota) and "pacing" (Worker-determined production rate), and Dutcher and his colleagues obliged."

Also, workers' regardless of how skilled they might be, were considered, at least by management, as 'only human' and that the work which was being patterned would be imperfect in some way, and performed more slowly than it could be done in terms of the cutting capacity of the tooling.

With the advent of NC technology, R/P was put on the shelf at GE, and stayed there to collect dust:

"Management liked numerical control better," Orrin Livingston, the consultant engineer who first thought of the phase-shift approach, recalled later. "It meant they could sit in their offices, write down what they wanted, and give it to someone and say, 'do it.'...With numerical control, there was no need to get your hands dirty or argue." Earl Troup, of the Industrial Applications Group, concurred. "With record-playback, the control of the machine remains with the machinist.... With numerical control, there is a shift of control to management. The control of the machine was placed in the hands of management - and why shouldn't we have control over it?"

Why not?, indeed. Management's perspective makes good sense, and I understand it perfectly. And the unions really didn't do much to help themselves either during this period. The belief in 'progress', having been well established culturally, along with a certain tendency towards technophilia on the part of those who operate machines, sloped the playing field for starters. Then, a decision within union management to position the unions as welcoming and supporting of technological advances, and to place union attention on contending with management in regards to hours of work, collective bargaining, grievances, wages, seniority, and so forth, ultimately allowed the unions to place themselves in a weak position, check and mate. The technology overran them, making vast numbers of workers obsolete.

At the United Auto Workers (UAW) convention of 1955, the union issued an unprecedented resolution on automation:

"The UAW-CIO welcomes automation, technological progress...We offer our a common search for policies and programs...that will ensure the greater technological progress will result in greater human progress. This goal will not be achieved, however, if we put our trust in luck or blind economic forces. We can be certain of recognizing the great promise for good and averting the dangers that would result from irresponsible use of the new technology only if we consciously and constructively plan to utilize automation for human betterment. We cannot afford to hypnotize ourselves into passivity with monotonous repetition of the comforting thought, that in the long run, the economy will adjust to labor displacement and disruption which could result from the Second Industrial Revolution as it did from the First."

Well, they did hypnotize themselves in a certain respect, which was that belief in 'progress', as something 'inevitable', even 'unstoppable', and something to which they could not be seen to be in opposition. The Luddites, by contrast, had suffered from no such inculcation of belief and when confronted with the threat of factory machine work dumbing down their trade and destroying their way of life, they responded in a direct manner by destroying the specific machines which accomplished that effect. Technology was not 'neutral' in their eyes, and they were themselves savvy users of technology. I've written about the Luddites in past posts at some length.

The trade unions of the post war period, in losing sight of the forest for the trees, if I can call it that, or ignoring the elephant in the room perhaps, focused their energies on matters which were really peripheral in import compared to the advent of automation. And with every strike, grievance and labor disruption, they gave management another reason to move towards the automation of production.

NC, and the later CNC which developed out of it, did not actually accomplish production cost reductions for the most part, but rather cemented control of production more firmly in the hands of management.

With more and more skilled workers being replaced or de-skilled by automated equipment, and little to no investment by big business in apprenticeship training, a new problem arises: a skilled labor shortage. The answer to this problem by industry? More automation. I'll look at some of the outcomes of this sea-change shift to automation in the next post, before returning to a more personal consideration of the role of CNC-equipment in the woodworking realm.

Thanks for coming by the Carpentry Way today.


  1. Chris, are you suggesting the significant difference is that RP relies on a skilled operator to do the original programing AND "playback", while NC programing and "playback" is set by programmers?

    Harlan Barnhart

  2. Harlan,

    no, not quite. Sorry if that wasn't clear. In R/P, like the player piano, a copy of a skilled machinist's work is first achieved by use of a stylus and recording device. This then allows for machines to be run -the playback phase- on the basis of that program copied from the machinist's work, without the machinist needing to operate the machine - at least not with the same degree of attention needed when the program was created. In the same way, a player piano plays the work of a skilled pianist, operating the keys by means of a mechanism, not human fingers.

    NC, on the other hand, is programming done by programmers, who often may have little understanding of machinist work, and work in an area isolated from the shop floor. Those programs are then used to run machines (playback), requiring only that workers load and unload, and otherwise watch over what is happening.

    I'll look over what I wrote to see if I can express it more clearly. Thanks for the question.


  3. One more thing - the stylus device mentioned above was only one tool for copying the machinists work, and was used for tracing the contours of shapes. There are manually operated milling machines and lathes which operate on the basis of stylus tracing which make no use of numerical control. One of the earliest mass production uses of a stylus tracer, for instance,developed in the 18th century, was for the reproduction of gunstocks.


  4. I see. Would you say RP is a technique used to replicate skilled labor and NC is a technology which takes creative skill off the production floor altogether?

    WWII is certainly a turning point in the story of American production. It seems to me the intense demand created by war destruction necessitated new systems simply because the old systems couldn't handle the demand. After the war, the systems stayed and consumer markets inflated to keep demand high. Perhaps the pre-war balance between consumption and production was more valuable than we imagined.

    Thanks for the provocative thinking.

    Harlan Barnhart

  5. Hi Harlan,

    "Would you say RP is a technique used to replicate skilled labor..."

    Yes, absolutely

    "... and NC is a technology which takes creative skill off the production floor altogether?"

    In theory, yes. It was often the case though that the programing didn't work as it was supposed to and machine operators, though they were not supposed to, figured out a certain amount of programming on their own to correct the production floor problems. So, that's about he remaining outlet for creative skill in a situation where design of the object and programming of machining to make that object are both done by non-machinists.

    "Perhaps the pre-war balance between consumption and production was more valuable than we imagined."

    This issue goes back to before the turn of the century actually- the 'problem of over-production' was well-established by the mid-to-late 1800's. If there was one thing that characterized manufacturing at the turn of the 1900's it was that it suffered from too much competition and too little standardization, both of which were seen by budding monopoly capitalists as problems to be solved. Industrial consolidation, mergers, and control of patents was was one set of means to limit competition, and product standardizations through the CSA and ASME, etc, allowed mass production to occur on a wider scale. The mechanisms were long in place before WWII, and things were certainly not all that 'balanced' in the years between the world wars, that much seems clear.


  6. Hi Chris –

    I have been enjoying this series, and this post in particular. I majored in Computer-integrated Manufacturing Technology in college. Basically industrial engineering with a little CAD/NC/CNC/Robotics and business (production planning, cost accounting, etc.) thrown in for good measure. I didn’t become consciously aware of fine hand craftsmanship until after I entered the workforce and happened upon a colleague’s issue of Fine Woodworking. That encounter started me on my path as a woodworker. I have long felt that there is a place for CNC in the shop of the lone craftsman, permitting the artisan to leverage the equipment to perform some of the more menial/repetitive tasks. Take for instance a project involving bent laminations pressed in a vacuum press. This sort of bent lamination typically requires a bending form of some sort. The forms require generating a pattern, cutting multiple identical ribs and then assembling these into the form. CNC is ideal for this sort of repetitive work – design the form part(s) in CAD, generate/download a tool path and rout out multiple, identical pieces from sheet stock. With a little foresight, one can even include features (such as a notch for a spline or dadoes) to facilitate part orientation and assembly of the form. I can also see the benefit processing sheet stock for products like kitchen cabinets, permitting the craftsman more time to focus adding detail on façade – face frames, doors and drawers – those aspects the customer touches/uses on a daily basis. Plus as a business owner there are other obvious benefits … Much cheaper with fewer hassles/headaches than hiring an employee … the machine doesn’t bitch, require vacation time, weekends off, etc.
    Best regards,

  7. Matt,

    thanks so much for your comment. I agree with what you say, and I will be touching upon some of the points you have raised, directly or indirectly, in an upcoming post in this thread.



All comments are moderated, so if you're planning to spam this, know now that your clicking and pasting is in vain. I do read the comments before posting, so your mission is doomed from the outset. All this time and effort trying to put your inane spam onto blogs -- is this how you want to spend your time on earth?

Please do me the courtesy of appending your name to your comment, even if posting under the 'anonymous' option. No name = deleted.

Comments NOT accepted include:

-those containing links unrelated to blog content
-spam of any kind, or ham for that matter
-did I mention that attempted spam postings will be non-starters?