|
DFMA a Multi – Functional Analysis Tool David Meeker Abstract Many companies use the DFMA methodology in there product development cycle, but their application of the tool is often limited to one specific task or small area of the entire Product design process. The paper will explore different applications of DFMA tool suite throughout the entire product design cycle. The application of design for service and design for the environment will not be covered but it is logical to see how application of these tools would cover more of the entire product development process. DFMA a Multi – Functional Analysis Tool Many companies start using the DFMA tool in its more traditional role of product simplification and cost reduction and although the sophistication and use of the tool increases it never really expands beyond that traditional role. It is possible to in fact use the tool through out the entire product development cycle, from early concept through end of life. Areas of use include but are not limited to;
Early Product Costing It is possible to do early product costing with DFMA. Taking past product cost data and looking at attributes of your particular product it should be possible to create a metric such as dollars per pound, dollars per cubic inch, dollars per axle, or dollars per horse power. In an over simplified generic example; lawn
tractors.
Next curve fit the data to come up with an equation that
best fits the data.
Early Product Costing at the Subassembly Level The next level of product costing is at the subassembly level. This helps to take into account changes that new products typically have a new feature change from the previous generation. In the case of the lawn tractor the above data was for 42 inch cut. If the next generation is a 44 or 46 Hp. inch cut than the above model is probably not a good predictor. Breaking the product into major sub assemblies and repeating the above process will enable you to great a more accurate cost as various subassemblies are improved or changed from the previous generation. There are other factors that can be incorporated such as raw material cost and product release dates. The Levels of Analysis Accuracy1 Level 1 or Parametric Level 2 or By Analogy Product feature cost models such as Cost per memory slot, Cost per Disk Drive Bay, Cost per cubic inch for the enclosure, and others as seen on other product family members. It then aggregates up all the feature costs to make a total estimate. – OR – Significant part cost analysis - On most designs there are a handful of parts that account for 80% of the total as‑designed cost. This method tries to identify the cost drivers, get estimates for those parts and use allowances for the rest of the system. The accuracy of this method is only about 5-15%, and it is directly related to the how much information you have managed to gather about the product to correctly characterize either features or major cost elements. The time to complete this analysis is on the order of 2-5 days. We most often use this form of estimate for marketing estimates or competitive bid estimates when full access to competitive hardware is not available. Level 3 or Engineered The accuracy of this method is only about 3-7%. We do fewer of these evaluations because of their cost in terms of money and resources, but they provide the anchor points to our experience. Product Benchmarking Definition - The continuous process of measuring products, services and practices against the best competitors or those recognized as industry leaders. 2 In 1979 Xerox wanted to understand, "how in the world could the Japanese manufacture [a copier] in Japan, ship it over to the United States, land it, sell it to a distributor who sells it over to the dealer who marks up the cost to the final customer, and the price the final customer pays is [still] about what it would cost us to build the machine in the first place".3 In the course of trying to understand the competition, Robert Camp of Xerox defined the benchmarking process. This required Xerox to look to the hardware as a source for understanding and comparison, a major part of the benchmarking process. Simply put Product Benchmarking is the act of measuring a product against some standard. In common industry parlance this is measuring your product against your closest competitor. Performance benchmarking is very commonplace in the computer industry, with several consortia and societies devoted to developing and maintaining common methods of test and comparison. 4 There are three major phases or steps in the product benchmarking process;
It is the analysis stage that the DFMA tool is used extensively to create cost estimates, Labor standards, Theoretical Minimum parts count to get some idea how good the design is and what VE/VA opportunities exist. Cost the Product We use the Boothroyd and Dewhurst (B&D) tools for sheet metal, injection molded, die cast and machined parts. The size and feature data taken during the physical teardown are fed into software along with the volume expectation from the literature search. Through our corporate purchasing department we have obtained representative labor and raw material rates from most fabrication locations that we use in the tools. Tooling type (e.g. hard, soft, ..) is usually determined by examination of the part and input into the B&D tool. If we can’t determine the tooling type visually, we make assumptions based on the expected volume. We report incremental piece part cost and tooling separately as parts of this corporation treat tooling cost as a program overhead cost and other parts of this corporation amortize tooling over the life of the build. For Printed Wiring Boards (PWB) we use an internally developed tool to estimate PWB costs. Previously we used to a fairly complex tool we bought from BPA, but found that supply cost variations drove the final price more than technical complexity factors. Assembly time is estimated using Boothroyd and Dewhurst DFMA® software, using a partially loaded labor rate. Test Labor is estimated using the test time of similarly complex internal products, again using a partially loaded labor rate. Develop Metrics We also take note of the date codes on major assemblies, and the stage manufacturing dates within the machine. This gives some idea of the inventory pipeline depth, and where inventory is being held – either in raw stock, finished goods, or at the outsource house. Vendor Quotes The DFMA tool has a widely know track record on providing
detail analysis of part estimates that can be used to
compare to vendor quotes. When there is a discrepancy
between a vendor quote and the estimate the DFMA tool
services a great vehicle to start a dialog with the vendor.
The critical part that many companies forget to do is to
take the learning from these discussions and feed it back
into the DFMA tool to make improvements in the data bases
and also feed the learning’s back to design engineering.
The ultimate evolution of this process can be found in the
vendors actually submitting there quotes using the DFMA
tool. Conclusion In conclusion the DFMA tool can be used many place through out the entire product development cycle to provide useful analysis aimed at speeding the product development process, and yielding a more manufacturable product at a lowering product cost. 1 Definitions were
created by personal conversation with Neil Albert, then President of SCEA a friend and former colleague. See
The Society of Cost Estimation & Analysis
http://users.erols.com/scea/ |
