How one fabricator transformed its high-product-mix assembly
April 30, 2013
Coordinating fabrication, welding, assembly, and painting within a high-mix, low-volume environment is neither new nor easy. The effort at one such fabricator demonstrates the commitment required, tools applied, lessons learned, and benefits realized when adopting cellularized flow manufacturing in a traditional, departmentalized organization.
After having spent the past 20 years serving as a lean contract consultant to organizations such as Nike, Lockheed Martin, Alberta Canada’s Ministry of Human Services, and a few hundred precision sheet metal and machine shops, I sometimes think I’ve seen it all. But once in a while I come across a company of unique character and culture, and I’m reminded that there truly are new things under the sun.
VSR Industries is just a stone’s throw from the Las Vegas strip. In 2008 the company wanted help cellularizing a formerly departmentalized process related to one of its best-selling products—metal slot machine bases for the gaming industry. If you have ever walked into a casino, you might think that the 6-foot-tall slot machines are rather large. In truth, the rather diminutive machines sit perched upon bases that can be 15 to 20 inches high, 18 to 34 in. wide, and of various depths.
Each hospitality management team attempts to provide their guests with a new and unique entertainment experience, and the slot machine base plays a subtle but important role. Customers can choose from different facades and trim, including a variety of bull nose and footrest styles. This requires potentially endless varieties of shapes, colors, and auxiliary pieces that VSR produces on demand and just-in-time. A new casino opening may call for thousands of new bases to be designed, scheduled, manufactured, assembled, and painted to meet critically choreographed delivery and setup on the gaming floor (see Figure 1).
VSR’s transformation from batch manufacturing to a paced-and-line, balanced production system has not been without challenges. If your organization is considering moving toward providing small-lot-size, real-time delivery without using a build-to-stock model, then you should find VSR’s story encouraging, enlightening, even inspiring.
VSR had attended improvement workshops organized by Nevada Industry Excellence (NIE), the state’s Manufacturing Extension Partnership, part of the network under the U.S. Department of Commerce’s National Institute of Standards and Technology (NIST) Manufacturing Extension Partnership program. Jeff Englehart, a project manager with NIE, visited the company and, in turn, introduced me to VSR’s management team.
My first experience with the fabricator was related to a setup reduction and line balancing project in its press brake department. The kaizen team also put together a value stream map to prepare for bringing welding and assembly into the manufacturing cell (see Figure 2). At the kaizen report meeting, Colton Vollmann, COO and co-owner of VSR, politely listened to the team’s recommendations, which included the demand for six press brakes.
Vollmann glanced up from his BlackBerry. “You can have four.”
I thought he must not have understood the data. I reiterated that even after our setup reduction effort, we actually needed 6.2 press brakes, but that with continued team efforts we would surely be able to achieve the average demand for 108 units per day with six machines.
This time he didn’t even look up from his BlackBerry. “You can have four!”
We were stunned. I am a press brake operator, and the math was not complicated. Physics simply demanded six machines. Our team retired to the kaizen war room, where after an extended period of silence, one team member stated in frustration, “The only way we can hope to perform the work with four machines is to completely eliminate setup.”
“What would that look like?” I asked.
“It would be the ugliest setup on the planet. It would look like a porcupine, with dissimilar tooling going every which way in every machine. No press brake operator in their right mind would want to run it.”
The team divided all the prints into four piles according to which hybrid setup could possibly be developed to produce the parts. In one day team leaders were able to design and build four new press brake setups that captured the nearly infinite number of parts required for this value stream. With just the simple addition or deletion of a length of press brake tooling, each 8-ft.-bed machine now contained all of the tools necessary to produce any part in the product mix without a changeover (see Figure 3).
This experience reinforced the fact that, when it comes to improvement, you can never assume that you’ve done all that can be done. And if Vollmann had not pushed us to the absolute edge of possibility, we probably would have accepted the setups for what they were and not uncovered a 33 percent improvement in productivity. The company also didn’t have to purchase two additional press brakes (not to mention staffing them), which likely saved the company close to $200,000.
My work there done, I moved on to work with other clients. Six months later Vollmann called me and said, “We need a tune-up. It’s not working like it was.” When I arrived, I found that the demand had increased from 108 units per day to 150, and the products had undergone some major design changes. This was not a tune-up; it was a major overhaul. In addition, the company wanted to incorporate final assembly into the process.
The challenge now was to sync assembly with fabrication and welding. Anyone who has manufactured and assembled parts in a make-to-order environment will understand the unique challenges of choreographing the confluence of multiple products—all high-variety, low-volume, complex assemblies.
We developed a line-side assembly process, moving a team of assembly personnel into the fabrication cell. This required developing a process that fostered flow, matching takt time (manufacturing rhythm) and balancing standard work (work assignments).
At 480 minutes of available time per day (450 after deducting break times), and with a demand of 150 units per day, the takt time required was 1 unit every 3 minutes. The entire line would have to be set to that pace. Assembly required more than 10 minutes of labor per unit. Therefore, 10 minutes of operator cycle time (OCT) divided by a 3-minute takt time resulted in a need for more than three people in assembly. We rounded up to four.
Each person would be assigned 2.5 minutes’ worth of work for every 3-minute takt cycle. Extra time would be used to ensure 5S activities were applied to maintain a clean and organized work area.
The current cell layout did not foster perfect flow. After a value-added/non-value-added observation, the team determined a value-added percentage of 70 percent. This meant that 30 percent of everyone’s time was spent not adding value.
The team developed a new layout that shortened travel distances between hardware insertion and press brakes, between the press brakes and assembly, and between welding and grinding. The reduced travel distance meant that eight people could perform work that previously required 12. (Note that no one was laid off due to this improvement.)
The team ran a daylong simulation and was able to maintain the 3-minute takt time, hitting all of its production targets with only one hour’s worth of work-in-process inventory in the pipeline. This represented an 87 percent WIP reduction—that is, 20 units in process instead of 150 (see Figure 4). Quality also improved. Because only 10 units were in process between the press brakes and assembly, workers could find problems quickly and provide instant feedback to the upstream process.
When asked what the greatest learning experience was, VSR Production Scheduler and Lean Coordinator Giancarlo DeGregorio said, “Understanding that you have to look at the entire process as a whole was critical for us. You just can’t turn up the volume over there and disregard the effect on the process down here. It all has to be timed and paced to work together. You can’t optimize one process and completely ignore how that will push unneeded material into the pipeline.”
It is kind of like moving marbles through a garden hose; one in, one out. You are kidding yourself to think that you are going to get more output by pushing the marbles in faster than they can come out. You’ll just force a jam-up.
VSR’s first model line has had remarkable results, and they have since been able to adjust the output levels to match new demands of 200, 300, and 400 per day while maintaining minimal WIP and maximizing efficiencies.
Over the past two years, the manufacturer has applied the same methodologies to its other product lines, including its wooden slot machine base manufacturing value stream. It also has adopted lean processes for its complex slot machine and kiosk manufacturing and assembly product families. Many products comprise more than 1,000 individual components (such as wire harnesses, circuit boards, and power supplies) that must be coordinated with not only internal fabrication proc-esses, but dozens of outside suppliers as well.
While purchasing from an independent source adds to the complexity of materials coordination for an in-house assembly process, the replenishment method is simple. In fact, regardless of whether the replenishment source comes from an upstream process on the shop floor or an outside supplier thousands of miles away, the replenish equation is the same: Divide replenishment time by takt time.
For example, for an internal process requiring a half day (240 minutes) to fully replenish a downstream process, divide 240 minutes by a 3-minute takt time. This results in a reorder point and a minimum stock of 80.
The same logic applies to outsourced components. If it takes 10 working days to obtain a specialized fastener from a source outside the company, the takt time is 1 unit every 3 minutes, and you have 480 minutes available per day, then the reorder point has to be every 1,600 units—or, 10 days × 480 minutes = 4,800 minutes / 3-minute takt time = 1,600.
This works perfectly—until, of course, a production cell doesn’t keep the predetermined pace. At VSR, here is where real-time production tracking, including visual or electronic kanban signals, plays a significant role.
As General Manager Dino DeGregorio described, “We have incorporated the use of andon lights utilizing light beams electronically tied to a scoreboard, which is programmed to match takt time. Once a product passes through the light barrier, the system counts and logs the production onto a visual scoreboard. Comparing these real-time results against predetermined takt time targets, the team [and anyone else] is able to confirm whether they are ahead, behind, or on pace with their production goals.”
In addition to traditional lean manufacturing tools like setup reduction, 5S (sort, straighten, shine, standardize, and sustain), and value stream mapping, VSR simplified and standardized many of the components. “Realizing that assembly is a large part of our business … we have taken steps through various departments [including design and engineering] to ensure that the product is assembly-friendly,” DeGregorio said.
In addition to higher morale, better teamwork, and a greater sense of ownership and accomplishment throughout the day, the team (and the company) has built more meaningful customer relationships by focusing on the customers’ need rather than whatever batch of material sits in front of a machine. The company has shortened its lead-time, which has shortened the time between paying and getting paid. This improves the cash flow of the company and reduces floor space previously covered with inventory.
Among many other accomplishments, VSR’s improvement efforts have resulted in:
VSR is setting the pace and leading by example, teaching not only its vendors but also its customers the value of adopting and adapting the lean production system throughout the entire enterprise. If you ever find yourself in Las Vegas (or any casino), look at the nearly endless variety of bases supporting gaming machines. The people at VSR have worked hard to design a manufacturing and assembly process that is capable of producing custom metal and wood cabinetry on demand, often within hours instead of weeks or months.
The experience at VSR proves that American manufacturing can have a bright future, but only if we show a similar hunger to challenge the status quo; demonstrate a willingness to try new concepts; foster environments that embrace change and demand continual learning; and find a way to benefit each other through the open exchange of ideas and experiences.