Let’s say that you could get results like this for a product:
- Labor costs: Reduced 42%
- Part count: Reduced 54%
- Assembly time: Reduced 60%
- Assembly cost: Reduced 45%
- Assembly tools: Reduced 73%
- Assembly operations: Reduced 53%
Would you be interested?
OK. What if there were a couple more factors thrown in:
- Product development cycle: Reduced 45%
- Total cost: Reduced 50%
Who wouldn’t be interested?
“I don’t know of any case where it is done that it doesn’t result in very dramatic benefits of cost reduction. It’s not ‘beat-up-your-supplier-type’ cost reduction, it’s an improved product that’s easy and less expensive to manufacture and assemble,” says John Gillian of a methodology that’s been developed. Given those results—which Gillian says are rather typical of what’s achieved, not some special cases—and given the efforts that are being made by OEMs and suppliers alike to cut costs and to hasten new products to the market, you’d think that this methodology would be more popular than ice water in Hell. (Not that we’re suggesting that there are necessarily any analogies between the Netherworld and the auto industry, although I’m sure that you can think of several points of commonality.)
In point of fact, the methodology that Gillian is talking about is something that you’ve probably heard of and possibly may be familiar with because it has been around since 1983. Gillian is the president of Boothroyd-Dewhurst, Inc. (Wakefield, RI; www.dfma.com); the methodology was developed and transformed into a software package by Professors Geoffrey Boothroyd and Peter Dewhurst...back in the early ‘80s. They received a National Medal of Technology for their work in 1991 from the first President Bush. Yes, design for assembly (DFA) has been around for quite a while.
But it seems as though two things have happened. One is that while many companies may have deployed the method, it has gone by the wayside. Even though if asked whether they do DFA or not they’d undoubtedly say, “Um, yes, ah, of course we do.” The second thing is that there have been a multitude of other approaches developed to improve productivity, which have become the flavors of the month. Consequently, things that have been around for decades are discarded. Gillian suggests that it would probably be beneficial to refer to DFA as “lean product development.” Sounds more au courant.
So for those of you who are not familiar with it, here’s what you need to know.
DFA looks at a total product. It is putting a product together and taking it apart and asking a series of questions while doing so. The answers to the questions lead to an assembly time and an assembly cost. One key approach is to consider the parts that make up the assembly in the context of:
- Does it move?
- Does it need to be a different material than other parts in the assembly? (E.g., if you have a wire, there needs to be the insulator and the conductor, two different materials.)
- Does it need to be separate piece for assembly purposes? (E.g., it may be that it is a box that contains other elements and that the top of the box needs to be removed so those other pieces can be put inside.)
“If it meets one of those criteria,” Gillian says, “then it is considered a necessary part. If it doesn’t, it should be considered for elimination from the product.” One has to think about this, it isn’t a sure thing. Fasteners and spacers are the sorts of things that partisans of DFA look at with great skepticism.
“If you combine those questions with the assembly time, then you can see where the high assembly times are, and if the parts don’t meet the criteria, then work to develop the design changes to eliminate the parts by combining their functions into other parts,” he says, adding, “The important part of it is that it gives you a specific and quantifiable way of looking at the product.” While this can be done manually, the software that has been developed provides a boost with regard to the analysis.
One of the problems as regards the implementation of DFA is that people consider the other element that’s often paired with it—DFM, or design for manufacturing. (In fact, you’ll note that the URL for the Boothroyd-Dewhurst website contains the DFMA acronym.) “A lot of companies that say they are doing design for assembly aren’t changing the design. They are doing minor tweaking. It’s almost more like manufacturing process improvements rather than making the sorts of changes to the design that could radically improve the process.”
Or, said more simply: People forget the design part and think only in the context of assembly. Manufacturing people use the tool for process development. But Gillian points out that it is important to involve designers and manufacturing people together.
Gillian points out that if you look at product cost, there are basically three elements. Part cost, labor and overhead. The labor cost—the assembly cost—is generally on the order of four to five percent. Yet people give that most of the attention. How many products are being “outsourced to low-cost countries” because the assemblers are getting paid a few bucks a month? But the part cost, he points out, is generally around 70%. “That’s what DFA attacks.”
“If you can simplify your product during design, you have a lot less to manufacture later,” Gillian points out, adding, “Lean manufacturing can result in single-digit percent improvements. If you improve early in the design stage, you’re talking double-digit percent improvements.”
Conceivably, not only would addressing part design predicated on functionality and assembly mean that it wouldn’t be necessary to send the assembly work overseas, but it could also permit additional invest-ment in the product (e.g., the use of higher-quality material) because of the reduction in cost achieved through the DFA analysis. That, in turn, could lead to more sales, improvement in market share, and...It may no longer be the trendy thing to do, but it surely provides some competitive advan-tages for those who actually do it right.