CHAPTER THIRTY-TWO
Why Automation Pays Off
AUTOMATION, WHEREVER INSTALLED IN A MANUFACTURING plant, has paid for itself within three years, and often much faster. And it has given the plant the ability to compete even against low-cost foreign producers. Yet in every seminar I have been running these past few years, a plant manager gets up and says, “I can’t convince my top management that automation pays.” And all the other manufacturing people in the group shout, “Amen!”
The payoff from automation is indeed both fast and high. But very little, if any, of it shows up in the measurements that most U.S. manufacturing plants, especially smaller ones, use. Our conventional accounting figures measure the costs of doing. The main benefit of automation lies, however, in eliminating—or at least in minimizing—the costs of not doing. To convince their top managements that automation pays, manufacturing people therefore will have to develop new and more appropriate measurements of costs.
“Good quality is cheap; it is poor quality that is expensive” has been a truism in quality control for fifty years. Yet few plants know how expensive poor quality really is. The accounting system assumes that the process works the way it is designed to work: that is, that it produces the desired quality, uniformly and throughout. A good many manufacturing plants measure the yield, the percentage of finished products that come up to acceptable quality standards. But that is only the top of the quality iceberg. It does not show where, in the process, things went wrong. It does not show how much time and money has already been spent in earlier stages to fix problems. It does not show how many pieces have been taken off the line and scrapped as substandard without ever reaching final inspection. These costs are hidden in the conventional accounting figures, in overtime and overhead, in scrappage, in overstaffing, and so on. Even in well-managed “high-quality” plants they often run as high as a third of total manufacturing costs, sometimes higher.
Automation builds quality standards and quality control into every step of the process. Equally important, it spots and advertises a deficiency in quality the moment it occurs and at the place where it occurs. And it does so at no extra cost. Wherever we have fully automated—in Japan; in the Erie, Pennsylvania, locomotive plant of General Electric; or in the Fiat engine plant in Termoli, Italy—the quality savings greatly outweigh the savings in payroll and manpower, maybe two or three times. In fact, quality savings alone are likely to repay the costs of automation within two or three years.
The second major economic benefit of automation also lies in a cost of not doing: the cost of “downtime” when production is being changed from one model to another. During that time when tools are cleaned, molds and dies removed and new ones put in, machine speeds reset, and so on, nothing moves. And in preautomated processes, downtime often runs to hours, if not to days. Conventional cost accounting does not, however, capture downtime; it is a cost of not doing. Cost accounting assumes that the line will produce the number of pieces for which it has been designed; if it is 80 an hour, then cost accounting figures that each piece costs 1/80th the line’s cost per hour. This works fine if there are very large runs of uniform products, say in the plant that produces half a million uniform radiator grilles for Chevrolets. But in the overwhelming majority of plants the runs are much shorter, models change constantly, and there is downtime and more downtime—during which people have to be paid, the plant has to be heated and lighted, and so on—all, however, without any production. In an automated process, production changes can be built into the process. As a result, downtime is cut to the bone and often eliminated. Again, the cost savings of cutting downtime usually exceed those in payroll and manpower and often pay for the entire investment in automation within a few years.
Finally, the reduction—or elimination—of downtime often gives a plant an entirely new capacity to generate revenues, in addition to cutting its costs sharply. It enables it to turn out an optimally profitable product mix and to serve more profitable markets. This can be particularly important for smaller and more specialized companies, for example, a maker of electric wiring devices or a small foundry. In the United States and Western Europe, automation is still seen as something for the “big boys.” But experience in Japan, where the government provides low-interest loans for small-plant automation, would indicate that automation is, above all, the most advantageous manufacturing system for the small plant. It gives it lower costs through quality and downtime savings and also gives it higher revenues through more profitable product mixes and market mixes.
Top managements and boards are right in refusing to make the large capital investments that automation requires simply because automation is “advanced.” It is not even enough to be told that American manufacturing industries (and those of any other developed country) will survive only if they automate—though it is true enough. Top managements are right in insisting on some measurement of the expected benefits, some estimate of the return on the investment in automation. But they are wrong—as are their manufacturing people who so loudly complain in management conferences and technical journals about the “conservatism” of their bosses—when they believe that these benefits are “intangible” and have to be taken on faith. The cost of poor quality and downtime to firms can actually be dug out of the available cost data fairly fast by any competent production account (and then verified by a cheap and quick sampling run).
Determining what additional revenues automation might make possible requires a little more: a conventional market study. But to estimate the benefits of automation also requires a change in the way most managements still look at the manufacturing process and its cost structure. It requires a change from a focus on the cost per piece produced to a focus on the total costs of the manufacturing process. But this, after all, is what accountants and accounting textbooks have been preaching and teaching for at least twenty-five years.
(1985)