Lean Design - Read an article on Lean Design

Published in the November 2005 Issue of Machine Design Magazine -

The Twenty Cost Levers of Lean Design

by Ron Mascitelli, PMP

One of the greatest frustrations that designers face is that most of the opportunities for product cost reduction occur during the early stages of the development process…when the specifics of a design are the least understood. Decisions made during initial concept development can inexorably fix many of the critical cost factors in a product, making it impossible to significantly reduce costs later on. Materials and processes are selected, suppliers are identified, complexity is determined, and synergy with other products is established (or essentially ignored, as is the case in many firms). Like it or not, the manufacturing cost of a product must be a primary design criterion from the very beginning, since most profit opportunities have already left the barn before the first drawings are created.
Unfortunately, there are significant obstacles to performing early cost tradeoffs. Sketchy design information makes any formal cost analysis inaccurate at best. Moreover, most design teams are a bit uncomfortable with financial stuff, and are often reluctant to dig in on cost before they have their technical act together. What is needed is a simple tradeoff tool that can allow designers to identify lower-cost design alternatives quickly and with reasonable accuracy. Well, help is on the way. The approach described in this article will provide your development team with valuable and timely cost insights, without needing to know a general ledger from a laundry list.

Building up Costs to Determine Profit Margin

        Before we can begin to make intelligent tradeoffs, we must have a clear understanding of what factors contribute to the manufacturing cost of a product. (Please keep in mind during the following “generic” discussion that every firm has its own cost accounting structure, so it is important that designers take the time to understand their specific situation.) We begin building up a product’s cost structure by starting with the most obvious costs: production materials and labor (often referred to as “direct materials and touch labor”). Since these costs are the most familiar, it is all too common for cost-reduction initiatives to stop here. Bad idea. Although materials and labor are critical considerations, they often represent only a fraction of the total cost buildup of a product, as shown in Figure 1. Items such as dedicated capital equipment and non-recurring design cost must also be considered. Unfortunately, many companies lump both capital investment and design costs into indirect overhead, thereby hiding them from view (and often from consideration). If a piece of equipment is needed only for a specific product, then that product should absorb all of its cost. If the equipment will be shared across multiple product lines on the other hand, it is not so easily allocated (more on this later). Likewise, non-recurring design costs should be directly assigned to individual products, and should be included in the profitability analysis for those products.
        Things get trickier, however, when we attempt to deal with the “non-assignable” portion of factory overhead. A firm’s indirect overhead is really just a big trash bin into which all costs other than assignable costs are tossed. Items such as maintenance, inventory carrying and handling costs, factory utilities, sustaining engineering, and so on, are all components of indirect overhead. Since (presumably) these costs are applicable to all products within a factory, they are typically allocated by some “logical” accounting scheme, most often as a multiplier on direct labor (and sometimes direct materials).
        Each of the major cost factors listed above can be thought of as a “knob” that can be turned up or down by product designers as they make their design decisions. Hence, to summarize, there are five “cost knobs” that can be adjusted by a development team: A) direct materials, B) direct labor, C) assignable capital equipment, D) assignable non-recurring design costs, and E) indirect overhead. A downward adjustment to any of these cost knobs will reduce the manufacturing cost of a product and increase its gross margin, provided that the other knobs are not negatively affected. Keep in mind that cost factors can often be connected to each other in subtle ways. For example, in the case of a high-volume product, it may actually make sense to increase assignable capital investment (e.g., by purchasing automation equipment), because the resulting drop in direct labor will more than make up for the increase.
        Okay, all this financial stuff is well and good, but what can a humble designer actually do to impact one or more of the high-level cost knobs described above? This is where our tradeoff tool comes in, but before I describe this practical technique, we need to expand our picture of product cost one more time. To give us the level of detail that we require, I will identify twenty “cost levers” that impact one or more of the knobs on our control panel, as shown in Figure 2. Each lever represents a possible tradeoff that designers might consider when attempting to meet their target cost (four levers per knob times five cost knobs). These levers will then be incorporated into a simple tradeoff evaluation tool that can enable rapid and reasonably accurate decisions on how best to squeeze the last nickel of waste from your new product.

Like Archimedes Said…It’s All About Leverage

To paraphrase the illustrious Greek mathematician, Archimedes, “If you give me a cost lever strong enough, I can move the balance sheet.” Cost levers are simply aspects of a product design that can have a significant impact on manufacturing cost. Before I share with you my choices for the twenty most important cost levers, I must put forward a caveat. The cost levers for your products and markets may well be different from the ones that I suggest. My recommendations are intended to illustrate the types of considerations that are critical, but you should modify the list to suit your specific needs. Now, in no particular order, are my top-twenty picks for design factors that can have significant leverage on product cost.

A. Cost Levers for Direct Labor

Lever A1: Simplify Manufacturing Processes – Can the production process be simplified by reducing the complexity of assembly, eliminating fasteners or interconnects, designing for top-down assembly, or by using standard tools, etc.? Can machine setup and changeover times be reduced?

Lever A2: Reduce Required Skill Level – It’s not just the number of direct labor hours that determines total labor cost; the necessary skill level required can have a major impact. By designing a product to be simple to assemble and test, and relatively easy to adjust or customize, the per-hour cost of labor can be significantly reduced.

Lever A3: Automate Manufacturing Processes – Automation is a wonderful thing, but it should be used only after careful consideration of the tradeoffs. Since capital depreciation costs are typically buried in the overhead rate, automation may seem like an easy route to lower labor costs and higher profits. In reality, a major capital investment taxes the profits of every product within a profit center, and depletes a firm’s reserve of available funds. Automation should be used intelligently, and only with a solid justification behind it.

Lever A4: Reduce Test / Inspection Requirements – Here’s a ripe opportunity for direct-labor reduction. In the strictest sense, test and inspection are non-value-added activities. If we can ensure that quality is maintained, any reduction in this area is just like printing money.

B. Cost Levers for Direct Materials

Lever B1: Reduce Scrap – In some industries such as semiconductor processing and precision die casting, the cost of scrap can be significant. In general, the scrap rate is driven by both the capability of the process (its ability to achieve the required tolerances) and by the robustness of the product design (its ability to accommodate process variability without a loss of quality).

Lever B2. Reduce Parts Count – One of the guiding principles of lean design is to “eliminate or standardize.” Can a part be combined with others and thereby be eliminated? If not, can it be made common with other parts in the product (e.g., can all fasteners in a product use the same part number)?

Lever B3: Use Cheaper Raw Materials / Parts – The opportunities here fall into two categories: 1) the material selected is an overshoot for the given application (e.g., using stainless steel when painted metal would do), or 2) expensive materials are being specified as a substitute for a more clever design (e.g., using high-precision electronic components instead of taking extra time to design a more tolerant circuit).

Lever B4: Use High-Volume Parts – This opportunity is related to parts-count reduction. If we can standardize on a small set of frequently used parts, then we save in two ways. First, the cost of material handling, purchasing, inventory management, etc., will be reduced (a favorable impact on indirect overhead). Second, the common parts will have a higher order quantity, resulting in volume discounts from suppliers.

C. Cost Levers for Assignable Capital

Lever C1: Eliminate Batch Processes – Oh, it’s a vicious cycle. First our firm decides that a new production line will require investment in capital equipment. Then someone looks at cost versus capacity and decides that bigger is better, and huge is better still. Now, of course, the hulking new piece of equipment requires long setup times, has five times the capacity currently needed, and requires a tower crane to move it. Although there are exceptions (which should be carefully justified), small, rapid-throughput equipment that is compatible with one-piece or few-piece flow is generally more economical, flexible, movable, etc.

Lever C2: Outsource Capital-Intensive Processes – Purchasing capital equipment is like getting a tattoo; it seems like such a good idea at the time, but the enthusiasm wears off a long time before the tattoo does. Acquiring your own capital equipment means: a) you will need to keep it utilized, thereby constraining many of your future design decisions, b) much hidden overhead will be spent to maintain it, c) you must employ people trained to use it, d) you’ve paid a high price to enter a new market…one which must be reimbursed before you see your first dollar of real profit.

Lever C3: Optimize Tooling Cost – One of the most powerful cost-saving concepts to come out of Japan, Inc. in recent years is the Toyota “Production Preparation Process (3P).” Within this methodology is a real gem; the “Seven-Alternatives” process. The idea is that for every significant cost item in a product, designers should consider the advantages and disadvantages of seven alternative manufacturing processes. This can be a mind-expanding experience, particularly if your designers tend to use the same few processes over and over again.

Lever C4: Avoid Dedicated Equipment – It’s scary enough investing in expensive capital equipment, but if that equipment will be dedicated to a single product, the risks are even greater. What if the product bombs? What if the equipment turns out to be more costly than expected to operate? Capital equipment should only be purchased in support of a core process capability; something your firm will be in the business of doing regardless of whether any single product thrives or dies.

D. Cost Levers for Non-Recurring Design

Lever D1: Reuse Existing Designs / Processes – Design reuse is just like printing money. You save non-recurring design cost, and get the product to market quicker besides.

Lever D2: Eliminate Unnecessary Complexity – One of the most exalted compliments one can pay designers is to tell them that their design is “elegant.” Elegance of design means that high performance, quality, and customer satisfaction is achieved in a remarkably simple way. This requires innovation, insight, an artful touch…along with a healthy desire to make money.

Lever D3: Avoid Gold Plating of Designs – Gold plating means overshooting the customer’s needs. If a car with four wheels is good, why not a car with six wheels? If a VCR with a remote control is good, why not clog it up with a hundred meaningless buttons? Customers will not pay for performance or features that overshoot their needs, but your firm will pay for the cost of including them. Better is the enemy of good enough, so focus on solving the customer’s problem, and keep the gold in your firm’s pockets.

Lever D4: Optimize Make vs. Buy – If you decide to buy a part instead of making it, you may not need to design it. Suppliers have smart engineers waiting by the phone to help you tailor their products to your needs. Typically, if the customization is within reason, their non-recurring design effort is free. In fact, many suppliers (particularly in mature, commoditized industries) will do a complete set of drawings for you before you even commit to an order.

E. Cost Levers for Operational Overhead

Lever E1: Avoid Major Changes to Factory Layout – I like to say that the most competitive companies in the world sell products that look like they’re customized for every single buyer, yet the factory can’t tell the difference among them. If existing workcells, flow lines, capital equipment, material-handling equipment, storage locations, and logistics support can be used, significant savings will result.

Lever E2: Reduce Raw Material / Work-in-Process Inventory – Inventory carrying costs can be a significant contributor to indirect overhead. Using Just-in-Time (JIT) inventory management is the key, but product designers must be thinking about JIT during the design process for it to work effectively in production.

Lever E3: Reduce Material-Handling Requirements – Material handling consumes labor hours, floor space, and in some cases, expensive capital equipment. Think about how the parts and raw materials for your new product will be handled. Are heavy subassemblies designed to be easily maneuvered? Are there crane hooks, handles, or other such features needed to reduce handling labor? Can large and cumbersome structures be designed in a modular fashion?

Lever E4: Reduce Use of Consumables – In some industries, consumables (those materials that are used up as part of the manufacturing process) can be quite expensive. Examples include wear on tools and cutting devices, lubricants, abrasives, glues, paints and finishes, etc. This is particularly important if your new product would be the only one in the factory to require a specific consumable.

The Twenty-Cost-Lever Tradeoff Tool

        At this point you might be thinking that this cost-reduction stuff is just too complicated. After all, it’s hard enough to get the performance and features right. Now you are being asked to consider an entirely different (and admittedly complex) set of tradeoffs. Five cost knobs, each with four levers. Move one of the levers in the right direction and presumably the product’s cost will go down…unless there is some nasty negative effect on the other cost levers. Herein lies the challenge. We are working in a twenty-variable space with potentially high degrees of interaction among the variables. Fortunately, the tool shown in Figure 3 provides a relatively simple means to evaluate these interactions, albeit in a semi-quantitative way.
The twenty cost levers are shown along the left-hand column of the figure. Each lever is provided with a space in the adjacent column for a “weighting factor.” This is simply a multiplicative factor that captures the difference in relative importance among the levers for your specific situation. For example, if you are considering a low-volume, high-value, complex product, your set of weighting factors would be very different from those of a simple, high-volume commodity. I typically limit my weighting-factor range to integers between 1 and 5. Any larger range and the tool’s output may be unrealistically skewed. For a first easy cut, set all the weighting factors to one and use the tool to give you qualitative (yet still very useful) insights.
        I’ve provided space in this sample template for three design options; you can, of course, expand the tool to include as many options as you wish. Normally the way I would use this tool is to begin with a “default” design option. Perhaps this is a concept that is most like your previous product designs. It might also be a “performance-optimized” design that has all the horsepower and features, but at an unacceptably high cost. Set all the “scores” for the Option 1 column that represents this default to zero. Now get your team together and conceptualize some alternative designs that will move one or more of the cost levers in a beneficial direction. Spend enough time discussing these options to gain a rough understanding of how they would look, what materials would be needed, how much capital equipment would be required, etc. Go through each of the twenty levers and decide how the new design option would compare to Option 1, the default option. If the new design is more favorable from a cost standpoint, give it a positive score (using a range of integers from -5 to +5). If the impact on a given cost lever would be negative, give it a negative score. The magnitude of the scores should be roughly proportional to the estimated impact. Naturally, just pulling numbers out of the air is not nearly as accurate as getting some real cost estimates for all options under consideration. Try at the very least to agree with your design team on the meaning of a -5 score or a + 5 score, and be as consistent as possible. Note that a score of zero means that the new design option is not significantly different from the default design for that particular cost lever.
        Once you’ve completed your scoring, multiply the weighting factors by the scores in each column to generate a total weighted score for each design option. A positive total score (relative to the default design option set to zero) indicates that the new alternative should have cost advantages over the default. A negative total score says that you are better off with the default design. Given the subjective nature of this tool, total scores that are close together (say, a difference of five points or less) are essentially a tie. If all weighting factors are set to one, the highest possible total score would be 100 and the lowest would be -100.
        The best thing about the twenty-cost-lever tradeoff tool is that all you need are a few rough design concepts to begin using it. Again, this is simply an indicator of promising new directions to pursue and a comparison tool for evaluating various possibilities. Give it a try on your next design. This technique can be the product-cost equivalent of Lasik surgery for your development team; their eyesight for cost-reduction opportunities will be much improved through its use.