Issue link: https://industrialengineer.epubxp.com/i/826827
June 2017 | ISE Magazine 41 Lean manufacturing presenters are fond of tossing out these questions – and answers: 1. What is the ideal number of pieces per lot? Answer: One. 2. What is the ideal number of pieces per container? Answer: One. 3. What is the ideal number of pieces per move? Answer: One. One more such question must be asked: What is the ideal number of pieces per machine? We would like to say, "one," but this answer immediately raises further questions: Do we mix only enough rubber in a banbury to manufacture one tire? Do we dip only one door handle at a time into a plating tank? Do we apply an oxide coating to only one silicon wafer at a time in a diffusion oven? Do we even consider single-cavity injection molding (plastics) when multicavity molding is state-of-the-art? From one point of view the answer to each question is yes because one-piece-at-a-time processing through all phases of production is a lean ideal. If we should make 10 blanks per hit on a blanking press, those 10 then would need to sit in a state of delay and be si- phoned off one at a time into the next stage of manufacture. This adds manufacturing lead-time, carrying costs and other ills of inventory retention. From another point of view the answer is no: Blanking 10 at once reduces processing time. A lean mandate is to reduce throughput time – in any way. It is true that most of the ele- ments of throughput time are in the general category of delay: We often speak, for example, of the 90 percent queue time phenomenon. As queue time and other delay elements are shrunk through vigorous use of lean techniques, the actual processing time looms larger. At some point in the transition to lean, the met- al-cutting time (or rubber-mixing, diffusion or plating time) itself becomes a candidate for compression. Multipiece pro- cessing, of course, compresses processing time. The issue has to do with whether to buy the big machine in the equipment supplier's catalog – or the medium one or the small one. In the supermarket we may get out the calculator app in our phone and figure whether the giant economy-size box really is cheaper per ounce. Occasionally we are surprised to discover that the maker is trying to pull a fast one, as we find the big box costs more per ounce. The manufacturing engineer who selects equipment rarely troubles to make the calculation. In making the decision about size of equipment, the dominant concern is to get a machine that will handle projected demand. Economies of scale Often it is possible to buy two or three small machines that have about the same capacity as one big one. Buying capacity this way has one strike against it, or, in some cases, two: 1. The big machine sometimes has quality and performance features that smaller models simply do not possess. A larger copier, for example, may be capable of much sharper images than a small copier, and perhaps the large one can blow up or reduce an original and the small one cannot. Since qual- ity should normally rank highest among factors to consider in selecting equipment, the good reasons for small machines sometimes are overridden by the quality capability of the larger model. The challenge to the machine-tool industry is to design high-quality little ones. We need them. 2. While quality is sometimes a factor, there is another issue that is always present: The conventional view that bigger is better, a view that is rooted in the notion of economies of scale. Do you have a home with a lawn that has to be mowed? If so, consider your reaction if your lawn-care re- tailer stocked a mower that cuts a 12-inch-wide swath and another that cuts a 24-inch swatch. Would you give any thought at all to buying two 12-inch mowers instead of one 24-inch mower? Surely not. For one thing, two mowers would take two operators. Labor costs being what they are, the two-mower option is unacceptable. In manufacturing, we have held to the same kinds of beliefs. The economy-of-scale concept is also known, among engi- neers, as the six-tenths rule: Double the capacity and the unit cost of output falls by six-tenths. Economy of multiples Now we are confronted with powerful counterarguments – arguments for seriously considering multiple small machines instead of one large one. A few of those arguments center around inflexibility, responses to downtime and the ability to react fast. Take inflexibility, for example. Big machines add capacity in big chunks. This leads to paying for capacity that is used only a few hours of the day. Months or years later, when de- mand has caught up, the machine is run day and night without time for proper maintenance, so that downtime eats into avail- able time and the machine dies at an early age. Why not try to add equipment in small increments the way we add labor? (Some automakers add whole engine plants this way.) Big machines are harder to install. Sometimes they have to be dug into pits or bolted to the structure. They may require special utility hookups, drains or noise baffles. Anchoring the machine to the structure saps the plant's flex- ibility. Next time the products change, the machines are in the wrong place, and they cannot be moved easily. L