If your Kaizen events have not significantly decreased your lead times to customers, you may want to consider your designs. Even if you successfully reduced your time in the shop, in a pure Engineered-To-Order (ETO) setting, much of total order cycle is often in frontend processes. First, let’s revisit my original opening comments from my post “Are You Still Quilting?”
Regardless of company size, most products enter the game as a single idea. And, then, the quilting begins. Over time, some products become standard offerings with a select number of models. In other cases, products will remain Engineered-To-Order (ETO) and the quilting continues. It is this latter group we’ll talk about today.
Typical comments related to ETO Manufacturers:
- We provide custom solutions to our clients.
- Our business engineers products to meet unique customer specifications.
- We don’t make more than 1 of anything.
- Every client wants something different.
- We distinguish ourselves from commodity suppliers.
I’m a big believer in naturally occurring patterns. If you are providing a product or service into a specific market, certain needs will recur and form dominant patterns. Design for Manufacturing & Assembly (DFMA) is all about determining the dominant patterns driving “80%” of the action. Once known, you can then go to work with DFMA tools to develop designs that provide late stage, mass customization of components, assemblies, & services.
Like Lean Manufacturing, the title is a bit misleading. It is more about creating highly configurable products with minimal unique designs, parts, & process steps. Each Lean tool you have in the box is designed to attack certain elements. Kaizen events target the 95% of the time no value is being added to an order resulting in incredible velocity. Single Minute Exchange of Dies (SMED) events focus on reducing the time equipment is down to change over from order to order resulting in exceptional agility & much lower inventories. DFMA events are focused on reducing activity drivers.
Specifically, there are 3 basic drivers DFMA is designed to dramatically reduce:
Number of Unique Designs – In a pure ETO environment where quilting is the norm, the opening count can almost be infinite. DFMA events initially focus on determining what the “80%” options/specs are. Once known, the team then focuses on coming up with designs that are very robust & configurable resulting in the ability to perform end of build customizations.
Number of Parts – For most plants, every individual part has a set of transactions attached to it that require human intervention of some kind. Purchasing requisitioning, PO create, receiving, ERP transactions, payables, material handling, inventory counting… Let’s not try to add all the steps if you then manufacture something with that item. Simply stated, part counts drive a lot of activity for most products. DFMA is focused on developing designs that use far less parts. For assemblies, hardware is typically a big part count. Consider self-registering mounting brackets that require few if any nuts, bolts, or washers.
Number of Manufacturing Steps – Most plants are still batch manufacturing & not yet doing single-piece flow, so each manufacturing step has an inbound queue & setup. DFMA teams develop designs that eliminate as many steps in the process as is practical. Think about a universal casting design with minimal machining steps versus several fabricated parts requiring weldments or hardware for example. If you’re already cast or forged, get closer to near net shape or consider additive manufacturing.
Goals — Once you know what product or service you want the teams to work on, it’s important to establish challenging, but achievable goals. Teams I have worked with over the years have achieved results as follows:
Reduce:
- # of unique designs by 50%+ & make them configurable.
- # of parts by 75%+
- # of steps by 75%+
- Footprint by 40%+
- Costs by 30%+
- Lead-times by 80%+
I realize that at first glance this may not seem possible. If you simply try to do a little bit less of what you’re currently doing, you are right! The faster your teams take out a clean sheet of paper & start fresh, the quicker they will come up with results along these lines.
Like most Lean tactics, DFMA is rooted in method & common sense. Once top-level goals are determined, it’s all about getting focused where it will count the most & then systematically applying tools to get to your objectives.

You may have to repeat the design steps until you achieve the targeted outcomes. It’s ok to send the teams back to the drawing board until they get that AHA! breakthrough.
Let’s briefly talk about a few what’s in it for you items!
Working Capital — There are only two dominant variables that drive an inventory investment – Customer Demand and Replenishment Lead-Time. If you’re able to compress time deeply, you will proportionately reduce inventory requirements. There are two parts to maximizing this benefit – Time & reduced safety stocks.
Let’s look at time reduction first. This may seem a bit counter-intuitive, but sometimes a minimal investment in semi-finished goods can buy a great deal of lead-time. This is particularly true if the designs of those items are universal or easily modified. For many companies, most of the working capital tie-up occurs towards the end of a products build:

Strategically positioning a supermarket with universal goods early in the value curve can buy a lot of calendar with a minimal investment. Of course, for purchased items, having a supplier stock the supermarket on a Vendor Managed Inventory (VMI) basis drops working capital needs a great deal more!
If you have gained velocity thru Kaizen & DFMA events, you won’t require
near the amount of Work in Process or Finished Goods inventories as they are
time based.

A reduction of safety stocks is also possible as you convert unique designs into more universal, configurable products. This is due to a pyramiding effect:

The individual item safety stocks tend to offset when their demands are combined into a universal product. On average, I see 40% to 50% reductions in required safety stock investments at a 1.64 service factor (95% Fill-Rate). For the real -world demands shown in the preceding graphic, a 42% reduction was achieved.

Soft Cost Reductions – For most companies I work with, I like to look at Accounts Payable (AP) data. In almost every case, when I Pareto AP data, I find that 50%+ of the transactions are tied to 5% of the money flowing thru the business. Simply stated, 50%+ of the team in these processes are focused on 5% of the money flowing through purchasing. I also use a similar approach for manufactured goods & get the same results!
I’m sure nobody designed products & processes to intentionally tie up over half the team to manage 5% of the money. By substantially reducing design & part counts, you can materially reduce your soft costs. You are also in a better position to negotiate VMI arrangements with your supply chain. Once in place, your teams can now be deployed on much more meaningful tasks or to take on growth with no additional folks.
Technology — Following a
successful DFMA in which the 80% case has been determined, it is relatively
straightforward to automate design & product configuration. Many firms have
made the leap from paper to CAD generated prints. Some have even moved to providing
shop packets digitally to the floor.

Some teams have moved to using automation tools for frontend
processes like Inventor’s Ilogic. I’m familiar with Inventor, so I’m using them
as an example… If you’re already using iLogic, it’s a relatively small step to move
to a cloud-based product configurator such as Inventor’s Configurator 360: https://www.autodesk.com/products/configurator-360/overview
For your “80%” case, your sales team can then perform product configuration
while at the customers’ site & then deploy directly into your systems.

There are so many benefits associated with this capability, I could write an entire post on this topic alone!
Summary – This is a big topic with a lot of elements. I thought about breaking it down into a few posts but decided to go for it… Look at your own situation & see if you have designed yourself into a corner that prohibits you from making great strides in becoming the company who sets market expectations. You will take share if you do!
Best Wishes for a successful 2019!
Mike



From this Table, if you have a 1-shift operation with 2 paid breaks, your “available” time is 2,300 minutes a week. If the total average units flowing through the line was 1,140 units, your Takt Time equals 2.02 minutes. Simply stated in single-piece terms, we need to have a line that produces a part every 2.02 minutes to satisfy customer’s rate of demand. The next element that we need to evaluate is whether the process sequences violate TAKT time.
It really doesn’t matter how fast the other sequences are. If you happen to be a fan of
Sequence “A” was redesigned to take 1.90 minutes and sequences “B” and “D” were combined to create sequence into “B”.
Let’s use a sport analogy to explain Cp. Picture a struggling field goal kicker, who is spraying the ball from sideline to sideline. In this case, there would be a very low Cp reflecting almost no control. Ask yourself if it matters all that much which way you pointed him. A leap to using Cpk is futile until you achieve acceptable control. To keep this posting brief, I’ve skipped over
Until you achieve numbers in this range, I would avoid looking at Cpk. Again, with no little or no process control, you’re probably not going to reach a sustainable gain. If you get to a 1.75 to 2.0 Cp that looks like this, it’s time to work adjustments to target Cpk. Pivoting the kicker will now have an impact.
I championed a 6-sigma project focused on extruded blank weights for high-dollar fluorocarbon compounds used to mold automotive parts. Since a low blank weight would result in unfills and scrap product, operators would simply produce heavier blanks. Remember our field goal kicker? Well, in this case, we told him that we would give him points for anything to the right of the left post…
It read, “Every morning in Africa, a gazelle wakes up. It knows it must run faster than the fastest lion or it will be killed.
For this writing, I’m not going to go into any of the ideas used to identify what you should be a stock item, how much safety stock it should have, etc… I want you to consider a classification tactic that ensures 80% of your best and brightest are managing 80% of the money flowing through your plant. The adjacent table reflects typical results for studies I have conducted in plants throughout the US.
Set up a spreadsheet with your customers down the vertical axis and the products they buy across the horizontal axis. Use revenue dollars for the data points. Sort total revenue by customer in descending order and sort total product revenue in descending order. Once you’ve sorted the data, then place heavy borders through the data where 80% occurs for both customers and products. You will see 4 quadrants: