Our Sites

Flow, quality, and operational excellence in manufacturing

A conversation with Kevin Duggan shows why good flow can’t happen without good quality

You can’t inspect quality into a product. It’s a well-known phrase, but say it to anyone outside manufacturing, and they’ll probably give you a strange look. Of course, these people didn’t spend a career focused on efficiency metrics and Henry Ford-esque division of labor, where production people produced, quality people inspected, and a defect was quality’s problem.

Manufacturers have turned to various improvement methods to break down barriers and introduce some common sense into the system. A QA inspector can’t ensure quality products if he doesn’t work with the operators who make the product; shop floor personnel can’t improve a product’s manufacturability if they don’t work with manufacturing engineering; a manufacturing engineer can’t make a product easier to manufacture if he doesn’t work with quoting personnel and the customer. And the customer is where it all starts.

This is Kevin Duggan’s message. President of North Kingstown, R.I.-based Institute for Operational Excellence, Duggan has worked with various companies on implementing a brand of improvement—operational excellence—that focuses on flow: flow of customer information in quoting and engineering; flow of work orders from engineering through programming and order release; and flow of those orders through the first operation to the customer’s receiving dock.

Flow and QA

As Duggan explained, “In the old way of thinking you’re thinking about machine utilization, material yield and scrap, and setup times. Flow isn’t in the equation.” In the operational excellence way of thinking, however, “the very first thing you think about is flow, and the second thing you think about is that quality enables flow.”

On the surface, nothing seems to hinder flow more than the quality department. An operator sends a part for first-article inspection. QA catches a problem and the job needs to be cut on the laser again. Flow stops. Wouldn’t good flow be so much easier without quality getting in the way?

This views quality as an independent entity, a police force on the lookout for defects. As Duggan explained, quality isn’t independent, of course, but instead intertwined with everything in the business. A quality part needs a quality manufacturing process, a quality quote, a quality materials supplier, and quality interaction with the customer. “It’s not just about making the right part, following inspection procedures, and ensuring the absence of defects,” he said.

Some fabricators have moved certain quality functions out of the QA department, placing inspection stations near workcells. But as Duggan explained, before changing anything, people should start where everything in the shop begins: the customer.

“If you look at quality in a silo and just think about how quality fits in flow, I say timeout. Quality doesn’t start there. It starts with the customer. How does the process work from the time a customer gives you a request for quote, including the specifications for the product? How do you make sure what you make is what the customer wants? How do you interpret quality data from the customer, and if the customer is coming up with the design, how does that design go through engineering and programming? There’s a whole set of activities that occur before you even get to where and how you should inspect a part.”

Because of customer inspection requirements (first article, etc.) and the nature of certain parts, some jobs do need to flow to the coordinate measuring machine (CMM), laser scanner, and other shared resources in the quality department. But does every attribute need to be checked this way? Is there a way to make the part whereby certain quality requirements can be met at the workstation itself with, say, certain jigs or fixtures? Is there a way to embed quality into the machine or process cycle itself? If not in the machine cycle, can quality be built into the process at a nearby station, such as with a simple go/no-go gauge?

Determining all this, Duggan said, starts by interpreting quality specifications the customer provides.

Kevin Duggan is president and founder of the Institute for Operational Excellence, an organization that promotes a brand of improvement that focuses on flow.

Flow Through Shared Resources

An often preached aspect of lean manufacturing states that products or product families ideally should flow in dedicated value streams full of those clean, lean workcells—with no shared resources. In the typical contract fabricator or job shop, though (and most other manufacturers, for that matter), you just can’t avoid sharing resources. For this reason, Duggan defines a value stream as something that doesn’t have to be processed through a specific set of machines (though it can be, if it makes sense).

Product families dedicated to a value stream simply share certain flow characteristics. One value stream might include all products that only need to be cut and formed; another value stream might include parts that are cut, bent, welded, and assembled.

To define value streams, Duggan suggests making a product-family matrix, a chart with products on one axis and processes on the other. He added that developing such a matrix that’s useful does take time, because it not only involves grouping products that share similar processes, but also similar cycle times and demand levels. The math can get complicated when a shop works on complex assemblies with dozens or even hundreds of components.

But he added that the complicated math done upfront can build the foundation for operational simplicity on the shop floor. The matrix does this by helping to identify value streams and what Duggan calls the “guaranteed turnaround time” within those value streams.

Every value stream has a range of products with a range of turnaround times, and the guaranteed turnaround time is tied to the upper end of that range, to the products that take the longest to produce.

“Those guaranteed turnaround times are created by establishing formal connections between each process in the family, not the standard one-piece flow from traditional lean,” Duggan said. “In job shops, FIFO [first-in first-out] lanes to connect processes are preferred over kanban systems … More advanced FIFO systems are needed to connect shared fabrication processes that produce a high variety of part numbers for the different families.”

Once established, though, these connections define what normal flow is through the value stream along with the resultant guaranteed turnaround time for the family. Anything that takes longer than the guaranteed turnaround time is identified as abnormal flow.

Sometimes grouping machines together makes sense, simply because it helps maintain flow—not because operators walk fewer steps, but because they can see which parts are coming next, so they know to retrieve the setup information and tools they need to be ready for the job.

Duggan added that if machines are shared resources or monuments that would be extremely disruptive and costly to move (and many processes in metal fabrication fall into these categories), they need not be moved close together just to decrease the number of steps operators take. A long walk between machines doesn’t necessarily disrupt flow, as long as the operator sees what’s coming next.

As an example, Duggan referred to a kanban-like system, but instead of being based on a certain inventory level, it is based on the guaranteed turnaround time. In one such arrangement, Duggan helped set up a system in which a material handler moved three pieces to the work-in-process location near the downstream operation every 15 minutes, an interval based on guaranteed turnaround time. Most important, this allowed the downstream operator to actually see what was coming next, get ready for the next job, and keep the flow going.

As Duggan put it, “The net result was creating flow through the shop, without the expense or disruption of moving machinery.”

Flow in QA

In many shops, laser arms, scanners, and CMMs in the QA department are very much a shared resource. To manage flow in these situations, Duggan has used what’s called sequenced FIFO lanes. Shared resources have trouble with conventional FIFO lanes. The QA room may have jobs coming in from all value streams at different times and rates, and a simple first-come, first-served arrangement can cause a serious traffic jam.

A sequenced FIFO ensures that every value stream is given equal treatment. A CMM with a sequenced FIFO might have five color-coded lanes, each signifying a value stream. The operator works through the parts in a specified sequence, handling one value stream, then another, then another. If every lane is filled to a certain point—signified by, say, a green outline—that tells people that QA is operating in normal flow, within its guaranteed turnaround time of, say, three hours. If parts sit in a yellow-outlined area in the FIFO line, the turnaround time may be a little more—say, four hours. Finally, parts in areas outlined in red signifies abnormal flow.

This arrangement shows a practical approach to Duggan’s operational excellence tagline: “Each and every employee can see the flow of value to the customer and fix that flow before it breaks down.”

When employees see parts sitting in the yellow squares, they know that flow is starting to break down, and so they make an effort to correct it. They just do it—no meetings or formal kaizens required. It’s what Duggan calls the “self-healing part of operational excellence.”

In this flow-focused arrangement, the roles and duties of people in the quality department change. Front-line QA technicians still inspect parts and address defects. “So what does the QA manager do?” Duggan asked. “His job isn’t to resolve the quality issues. His job is to set up processes that resolve quality issues. There’s a big difference. Put another way, the QA manager does detailed work about how information is sent, to make sure no clarifications are needed. He needs to set up the way information flows.”

The QA manager works with others to set up a procedure so that people have the right information at a specified time. For instance, at a certain time of day—say, at 11 a.m.—managers walk to the quality department to retrieve worksheets with information about defects, and by 3 p.m. they come back with answers.

QA techs and others have filled out all the information on the worksheet, so no one needs to nag people for information, and there’s no need for an impromptu meeting. With the information on hand, personnel can go about their business to resolve the quality problem: Was there a material problem? An error in setup or fabrication procedures? Was a part signed off after an earlier process without the proper inspection?

That four-hour time period is the “guaranteed turnaround time” for a QA problem. Anything that’s addressed within that time frame is part of normal flow. The turnaround time also happens at a regular, scheduled interval, so people can plan for it and work it into their daily schedules.

This normal flow process in QA (see No. 6 in the sidebar) is designed to address the abnormal flow in production (No. 5 in the sidebar) that the quality problem created. If it takes longer than the guaranteed turnaround time for QA to get an answer about a problem, the QA process itself has abnormal flow that needs to be addressed. The more effectively information flows, the better QA can work with the rest of the company to ensure the work keeps flowing.

This pretty much boils down Duggan’s message. Manufacturing excellence doesn’t hinge solely on a workcell’s specific uptime or a lightning-fast setup. From a broad perspective, it relates to flow, both of the products themselves and the information about those products.

Kevin Duggan is the author of Design for Operational Excellence: A Breakthrough Strategy for Business Growth, and Creating Mixed-Model Value Streams, among other books, and the founder of the Institute for Operational Excellence, 1130 Ten Rod Road, Suite A202, North Kingstown, RI 02852, 401-667-0117, www.instituteopex.org

Tackling Offense

Manufacturers use myriad improvement methods, from lean manufacturing rooted in the Toyota Production System to the theory of constraints, which focuses on managing throughput around a constraint process. They all attempt to shorten the quote-to-cash cycle and uncover additional capacity—ideally without adding machines or people.

Because the manufacturer has a shorter quote-to-cash cycle than its competitors, its costs are lower, and so it can quote competitively on new business and still maintain or even increase margins. The business has more cash, too, because less of it is tied up in inventory.

That’s great for cash flow, but to realize greater profits, the business still needs to fill at least some of that recently discovered capacity. Otherwise, the company is writing the same (or similar) checks for materials and labor while bringing in the same amount of revenue. Salespeople can bring in more work, but the sales team has finite capacity too.

Here’s where playing offense comes into play: discovering new markets and new opportunities. Kevin Duggan, president of North Kingstown, R.I.-based Institute for Operational Excellence, organizes his approach to improvement so that managers need not concern themselves with the day-to-day operations of the business, but instead “work offense” to grow the company. Although this doesn’t happen overnight, Duggan said that once operational excellence is implemented fully, the shop just runs, and employees conduct and document improvements on their own.

Duggan outlines the process in eight steps: (1) Design lean value streams; (2) make lean value streams visual; (3) make flow visual; (4) create standard work for flow; (5) make abnormal flow visual; (6) create standard work for abnormal work flow; (7) have employees in the flow improve the flow; (8) perform offense activities.

Employees know where the work is (No. 1); see how it flows, know what’s coming next, and know how to process it (Nos. 2-4); can identify what’s not flowing as it should (No. 5); and ultimately know how to deal with that abnormal flow and improve existing flow (Nos. 6 and 7). This leaves managers free to grow the enterprise (No. 8)—to work “on” instead of “in” the business.

About the Author
The Fabricator

Tim Heston

Senior Editor

2135 Point Blvd

Elgin, IL 60123

815-381-1314

Tim Heston, The Fabricator's senior editor, has covered the metal fabrication industry since 1998, starting his career at the American Welding Society's Welding Journal. Since then he has covered the full range of metal fabrication processes, from stamping, bending, and cutting to grinding and polishing. He joined The Fabricator's staff in October 2007.