The recession forced shops to decrease batch sizes, and the habit may be sticking
August 1, 2011
The long-held belief that setups should be amortized over the largest possible lot sizes needs to change. As setup times become shorter, they become less relevant, and, according to some, so does the old notion of economic order quantities.
At one time Gary Conner spent much of his workday operating a laser cutting machine. Today the consultant at Oregon-based Lean Enterprise Training has a Shingo Prize under his belt, an award that recognized his book, Lean Manufacturing for the Small Shop. In it he wrote something profound:
For years, the idea of buying in bulk at a discount has appealed to the frugal homemaker and purchasing agent alike. Bulk warehouse stores are a testimony to the “bigger is cheaper” mentality. Companies need to re-educate customers. This requires communicating that the ordered product can be made in economical lot sizes based on the customer’s immediate need, rather than on some super-sized truckload … Customers have been trained (by suppliers) to believe that buying large batch sizes will save them money. The root of this evil is exorbitant setup times. The long-held belief that “bigger is better” is propagated on the idea that companies must amortize the cost of setups over the largest lot size possible. This idea needs to change.
The book discusses the economic order quantity (EOQ) formula, traditionally used to determine optimum lot size: too large and you get excessive storage costs; too small and you have skyrocketing setup costs. As Conner wrote, “EOQ seems like a logical approach. But the problem is that it assumes setup costs are fixed … If you reduce setup time to the point of it being insignificant, then the EOQ formula can be a waste of time.”
The book’s second edition was written in 2008 and published by the Society of Manufacturing Engineers in early 2009—not a stellar time for manufacturing. Oddly enough, the very nature of the recession and economic recovery may have been manufacturing’s saving grace. Consumption didn’t come roaring back as it did after previous downturns. Lackluster consumption combined with credit market troubles forced companies to order just what they needed immediately, and no more.
Suppliers who successfully cut their setup times could bid competitively on numerous small jobs. In spring 2009, Eric Hill, president of Cincinnati Laser Cutting, a job shop, said it was like “catching krill.” Like never before, the shop was full of that krill—numerous small jobs, and like whales, the shop had to gulp down plenty to keep growing.
Eliyahu Goldratt, physicist turned business consultant and the guru of the theory of constraints (TOC), noted the trend during a speech in early 2009. Tight credit and profound uncertainty were forcing everybody in the supply chain to lower inventories and order just what they needed for the immediate future. In other words, they ordered less, more often. Of course, during the worst of the recession it wasn’t as often as company managers would have liked.
In two of Goldratt’s business novels on TOC, The Goal and It’s Not Luck, companies won work not by outproducing rivals and cutting prices, but by changing over jobs quickly and shipping just what customers needed when they needed it. One big shipment turned into numerous small shipments and, ultimately, greater manufacturing profits.
The recession essentially forced everyone to abide by improvement methodologies like lean and TOC at least in one respect: They reduced batch sizes and, by necessity, setup time. With customers ordering only 10 of this, 15 of that, and so on, fabricators had no choice.
As these improvement techniques preach, material movement is what really matters, and to move many small lots through the shop quickly requires frequent setups. Now that manufacturing is leading the recovery, the industry may well be hiring more people who can perform those setups. The industry needs programmers who know how to nest multiple small jobs on the cutting center. It needs press brake personnel who know how to set up both new and (ideally) old machines. It needs welders who can join various products. And the industry needs engineers and designers who know how to make shop floor operations easier by looking at a drawing, working with the customer, and making a few design-for-manufacturability (DFM) changes.
Mainstream media stories often attribute manufacturing’s skilled-worker conundrum to automation. Only highly trained people can run such sophisticated equipment, they say. That’s true, but it may be only half the story. Old iron lasts.
Consider hydraulic press brakes, equipment that takes high operator skill. The latest “Machine Tool Inventory” study, published in 2008 by the Fabricators & Manufacturers Association Intl., surveyed about 1,600 fabricators and produced telling results. Respondents reported having 313 hydraulic press brakes less than 4 years old; 407 machines 5 to 9 years old; 382 brakes 10 to 14 years old; 209 machines 15 to 19 years old; and 277 that are 20 years old or more.
The press brake is a relatively labor-intensive machine, compared with modern punch presses and laser cutting machines, and the ages of those hydraulic brakes span a wide range. Then there are the 1,340 mechanical press brakes that respondents said were more than 20 years old. I’ve heard fabricators refer to these hunks of iron—affectionately—as “dinosaurs.”
Although machines have gotten more sophisticated, older machines often require more skill to use, and thousands remain on shop floors. So when it comes to the skilled-labor shortage, sophisticated machinery may play a role, but the real challenge perhaps has been with batch sizes. Skilled people interact with equipment—new and old— more often because batch sizes are so small and setups, though extremely fast (especially on new equipment), are more frequent.
It may be true that setups on new equipment are so fast and simple that they mitigate the need for highly skilled workers. Thanks to offline programming, even a new operator can set up a machine: The program tells him everything he needs to know, and the brake often shows the operator exactly where the punch segments should go. Setup isn’t so hard, right?
But somebody has to write or verify the program. The highly skilled person may spend more time in front of a computer screen, monitoring bend sequence simulations, ensuring blank sizes are correct for available tooling, and so on. And thanks to small batch sizes, he’s monitoring the programming of many different parts each day. That takes skill and knowledge. Besides, to get the most out of all equipment, shop personnel should know how to run those older brakes too.
In a sense, small batch sizes may be defining modern manufacturing in small job shops, which comprise the largest segment of U.S. manufacturing. In fact, more than 60 percent of the 1,600 metal fabricators who responded to the “Machine Tool Inventory” survey work for companies with 50 or fewer employees. In 2002 Census Bureau statistics revealed just how many small companies there are in manufacturing: More than 90 percent of firms have fewer than 50 employees, and more than 40 percent have four or fewer.
Perhaps one day, as shops continue to reduce setup time, the old EOQ formula will fall by the wayside. Material costs, of course, will continue to be a factor, but setup costs may play a smaller role. If job shops continue to reduce setup times to the point of insignificance, their customers’ purchasing managers may need to relearn their jobs. Buying in bulk may not be necessarily better or, for that matter, cheaper.