When Charlie Chaplin, in the movie ''Modern Times,'' scratches his nose and loses his place on the assembly line, his doomed struggle to catch up etches the definitive image of man held leashed by the machines of his production.
It's a relationship that bears watching.
Nothing is quite as symbolic of the benefits of mass production as the automobile. Yet the sometimes lockstep process of putting cars together alternates between a symbol of low-cost efficiency and the road to the blue-collar blues.
Now, with Detroit often typecast as the ''heavy'' in discussion of the malaise of American heavy industry, the assembly line is being thrown into turmoil.
Computerization is forcing a quantum leap in manufacturing technology, presumably leading to cheaper, better cars, assembled with far fewer people.
As the great comic Chaplin assembled cars with a wrench in each hand, today's auto worker faces a TV-like screen emitting a ghostly green glow. The screen is an output terminal for computer-aided design (CAD) which, along with its sister, computer-aided manufacturing (CAM), is high tech come to Detroit.
The tubes have changed things. Not the tubes exactly, of course. Like the engineer's slide rule and chalkboard, together with the skilled machinist's trade, the tubes are both tools and symbols.
CAD-CAM is both process and a machine, embodying half a century of struggle to apply unwieldy mathematics to unpredictable automotive design and manufacturing.
The underlying principle of CAD is to take a three-dimensional design, such as the shape of an auto fender, trace it, and store the image in a computer memory in the form of an almost infinite number of mathematical points in space.
The image can later be retrieved by ''playing it back,'' displaying its image on a TV-like screen.
Its value to engineers lies with its potential for storing other information as well: weight or physical properties of steel, for instance. The most powerful systems allow each part of the image on the screen to react to stress, heat, or vibration, much as would a real part.
Coordinates taken from the image can run numerically controlled machine tools , actually carving out the physical object - or its mirror image in a production die.
While it originated in the aerospace industry, CAD has been adopted by automotive engineers, who, in the past, coped with design by a kind of cut-and-fit methodology. In fact, the environment encountered by the family car is a trickier design problem than outer space: temperature extremes from 30 below zero to 110 above, potholes, even crashes.
It is the kind of stress that defies neat analysis.
Automotive engineers still don't know everything they can do with CAD. They know, for example, they can do stress analysis, impossible only a few years ago; get accurate thermal-distribution patterns; develop thinner castings.
Too, they can even take a picture of a car on a screen, send it screaming toward an imaginary barrier, and tell how much it would have crumpled if it were real. Chrysler Corporation has programmed an imaginary man who sustains imaginary injuries.
Ultimately, the effect on mass production of the finished design is expected to be even more profound. The other transitions, from idea to drawing, to tooling, to machining line, are being cut short, altered, even eliminated.
Designs done ''electronically'' can be translated directly into machine-tool language.
Coded information then can theoretically be transmitted to run machine tools anywhere in the world. It's still Buck Rogers stuff, but the prospect looms of a central design staff with the ability to set up manufacturing plants anywhere with far less hand work than in the past.
Even while engineers try to figure out what they can do with CAD-CAM, they really don't know what the new technology is doing to them.
With all its power to increase efficiency, CAD-CAM can potentially also eliminate thousands of engineering man-hours, translating to fewer jobs, even for highly trained people.
The impact on the production line can be even greater in terms of people affected. CAD-CAM demands new skills and puts a premium on types of education long foreign to the automobile industry. It demands new schools, new people, and new ways of looking at old problems.
Today's engineer never gets very far from the classroom, and the shortage of skilled operators has already prompted auto companies to set up independent CAD-CAM training centers of their own.
Meanwhile, the new technology is quickly forcing ethical judgments on how far to go with technology in a job-hungry economy awash with unsold cars.
At the crux of the CAD-CAM revolution is the emerging multinational auto industry. Unwilling to join the list of industries that today are dominated by offshore manufacturers - including textiles, shipbuilding, and consumer electronics - US-based auto companies are in a head-to-head race to improve productivity, and thus reduce the cost of their products.
The high-tech revolution is part of a massive retooling program that likewise includes reducing the size and fuel consumption of domestic cars and adopting new styles of management.
The goal is a more flexible kind of process that makes changing new-car design both quicker and cheaper.
''We did the 1975 Seville in just 13 months - the first time we tried a whole project,'' says Robert J. Templin, chief engineer for Cadillac and a CAD pioneer.
''Everybody was asking us: Did you really do that, or did you just luck out?'' he adds. The amazement of his peers stemmed from past experience, up to three years of development time shortened by two-thirds.
If there is irony in this move to computerized manufacturing techniques, it is that the high cost of introducing such systems is frequently blamed for the current high price of automobiles.
In the midst of an $80 billion program to renovate an industry that did pretty much ''business as usual'' from the end of World War II to the mid-1970s, the expense of developing sophisticated techniques increases costs today, the gamble being that costs will be reduced in the future.
Moreover, the cost reduction anticipated is mainly the ability to do with fewer workers - and the inevitable secondary effect on the economy stemming from unemployment.
The specter of a workerless, hence customerless, economy looms in any discussion of automation.
''Generally, productivity is considered on one side and the social impact is a separate discussion,'' says Harley Shaiken, research associate in science, technology, and society at the Massachusetts Institute of Technology. Mr. Shaiken notes that General Motors commissioned a study of the social impact of CAD-CAM in the late 1970s and then backed off.
''The scope of this process [CAD-CAM] affects everything,'' he says. ''It presents serious questions about employment. It is very vital to be competitive, but that doesn't address what happens to the displaced.
''The danger is where innovation translates directly to unemployment. You end up with tremendous resistance to technology. You have to get beyond that. A rapidly expanding market historically was a key mechanism to account for automation. Now one of the key cushions is gone.''