How cellular manufacturing can work in contract manufacturing
How a fabricator revamped its floor layout, and why it hasn’t looked back
For the past year one contract fabricator in Georgia has operated under the cellular manufacturing concept. Even in the shop’s high-product-mix environment, the concept is working—even thriving.
In one area of Metcam’s shop floor sits a punch press that feeds parts to a robotic press brake, which feeds parts to an area with a hardware insertion press, and manual and robotic gas metal arc welding machines. Workers produce enclosures for a unit that converts the direct current of solar arrays into alternating current. Metcam won that large contract earlier this year, and part of what helped the Alpharetta, Ga.-based fabricator was that cell concept, built to produce the product family (see Figures 1 and 2).
Last year this corner of the shop looked completely different. Back then workers assembled components for a large electrical components company. Metcam still does that work, just in a different place. Earlier this year that work had to be moved to another building off-site. But over the months the company reduced work-in-process (WIP) in another area of the plant, allowing it to move the operation back into the building, transforming inventory space into productive space.
Meanwhile, the company had another bottleneck to overcome—the powder coat line. So the shop installed another paint line, but with an unusual parts washing system, a modular one that can be assembled and disassembled over a few days.
All this change occurred not over years, but months. Adapting cellular manufacturing to the job shop can work wonders when it comes to productivity, scheduling, and on-time delivery. It also requires change—lots of it. But according to managers at Metcam, change is not a bad thing. It’s what continuous improvement is all about.
A Lot of Racking Going On
In February of last year The FABRICATOR covered how the shop rearranged its shop floor into cells, each with machines for cutting, bending and forming, hardware insertion, and other processes. The argument against manufacturing cells in contract metal fabrication has always hinged on product mix. Historically, manufacturers developed cells that were product-specific. How could cells possibly work for an operation with dozens, hundreds, even thousands of different part numbers?
Metcam’s managers had a different perspective. They discovered that for their diverse part mix, every part was cut, most parts were bent, and most also required hardware. They also found that separating these processes by department actually created tremendous amounts of WIP. The cutting department cut blanks and put them on racks. The bending department retrieved those blanks from the racks, bent them, and then put them back on yet another rack. Workers processed the part, racked it, processed it, and racked it.
So the company set up cells (initially seven, now eight) designed primarily around—but not exclusively for—product families (see Figure 3). Most jobs flowing through a particular cell may be part of the same product family, but not all jobs. Certain hot jobs or short runs may be assigned to the same cell, scheduled in between the longer runs.
According to Jerry Ward, vice president of operations, the arrangement promotes high-velocity production. Parts may be cut, deburred, bent, and have hardware inserted in the morning; welded in the afternoon and painted at night; then shipped to the customer the next day.
Ward added that flexibility, including the ability to squeeze in unexpected work, is one of the primary benefits of cellular manufacturing. In the past, if a hot job came in the door, the scheduler had to squeeze in the job between batches at every process: laser cutting or punching, bending, hardware insertion, joining, and assembly. All that squeezing is effectively what made the hot job “hot.”
Cells have made it so the first manufacturing step is what used to be multiple processes: say, the “punching-bending-hardware” step. Instead of negotiating time with multiple departments, the scheduler now makes that hot job the next order in queue. Hot jobs are no longer “hot.” They’re just, well, the next job.
This significantly increased available capacity in those primary operations, so much so that managers soon realized they had another bottleneck to contend with, this time at the powder coat line. So the company added a second finishing line (see Figure 4), installed by Conyers, Ga.-based Automotive Industrial Solutions (AIS). The powder coat line itself came from Somerville, Ala.-based Reliant Finishing Systems.
But the decision wasn’t that simple. In cellular manufacturing, parts move through multiple processes quickly, spending little if any time as WIP. But a powder coat line is at its core a batch-style process. Because of all the plumbing, not to mention the drying furnaces required, it just doesn’t make sense to have one painting or coating operation for each cell.
Still, changing over a paint line isn’t too arduous, especially if you aren’t reclaiming powder, something most fabricators don’t bother with anyway, because the time and effort that it takes to reclaim the powder costs more than what fabricators save on it. This sometimes applies even for product-line manufacturing. Ward described how one of Metcam’s largest customers invested in a powder reclamation system and ended up not using it. “It just takes too much time,” he said, “and time is so important.”
Technicians ensure parts for the next job are hung and staged at the ready, to minimize downtime between runs (see Figure 5).
Then, during the paint changeover, they simply leave a gap in the line, during which they have about five minutes to change out their equipment. “Generally, we have one or two people helping to prep the changeover,” Ward said. “We have backup guns, so that the paint technician can just roll right in and start with another color.”
Although the company changes out colors as needed, it does hold a little WIP before the paint line. That work mostly involves special colors. “One of our customers has products that come in 20 different colors,” Ward said.
A few colors dominate, while the remaining “specials” involve extremely short runs. If the company actually processed those parts as soon as they arrived in the paint staging area, the paint line would need to change over more than a dozen times a day, which would put a serious drain on throughput.
What makes powder coating a little inflexible is the pretreatment that comes before it. Some parts can require four or five stages of pretreatment others may require two or three. Moreover, the space between those metal prep stages may differ depending on the part geometry. Long parts require a lot of space between stages; short parts, not so much.
That’s why most shops install a preparation line large enough to accept its largest parts and then some. Metcam’s second powder coat area has a pretreatment line that’s 93 feet long, long enough to accept its largest parts (see Figures 6 and 7).
The system was developed and first used by Mankato, Minn.-based Associated Finishing, Inc., a custom coater that was instrumental in helping Pretreatment Equipment Manufacturing (PEM) Corp. design and produce the SprayLean product. The inline pretreatment system consists of modular 4- and 6-ft. stages, made of a stainless steel frame with a poly tank and shell. The stages can be installed in a certain configuration, then expanded, contracted, or otherwise altered if the shop’s metal prep needs should change.
“Each module operates independently,” said Ted Schreyer, vice president of PEM. “So say you start out and just need a two-stage washer; maybe you’re just cleaning and rinsing. Now you want to go to a three-stage process; all you have to do is add more modules. It all locks together, almost like LEGOs.”
Metcam purchased a pretreatment line long enough to handle its largest parts. If the line happens to be cleaning parts that don’t need so many stages, the technician simply turns off a few spray modules—a typical practice. But what if someday Metcam’s product mix changes to the point that it needs a third paint line for smaller parts? The fabricator now has the option of removing some modules in its current line and using them to build that third line. The company has no plans of doing this anytime soon, but change is inevitable in contract fabrication. Shop floor configurations transform with the product mix, and this someday may include changes to the painting operation.
“We work off of some of the ideas behind the theory of constraints here,” Ward said.
Several years ago, before the shop’s transformation to a cellular layout, WIP was everywhere. Machines ran flat out all the time, but fewer products were shipping out the door. Parts didn’t flow, but instead slogged forward from one rack to the next. Now those racks are gone, and the little WIP that remains makes the constraint processes stand out.
This includes the WIP in front of the paint lines (see Figure 8). The vast majority of work ahead of the paint line won’t be there long. “Most of this will be painted this afternoon,” said Ward. “And it will be gone tomorrow morning.
“I don’t tell anyone [in upstream processes] to slow down because the paint line is behind,” Ward continued. “I want everyone to run wide open, and I’ll figure out how to get it through.” This may include moving certain people to the paint prehang area or to masking.
“I see these guys mess with each other, too,” Ward said. “Every time the paint backlog gets a little low, the people in the paint area say, ‘Bring it on.’ And the people upstream ramp it up and get parts to them. And they say, ‘OK, big boy. You’ve got your parts now.’ They have fun with it.”
The fabricator’s paint lines adapt a traditionally large-batch process to accept the small lot sizes received from upstream cells. Now the company is tackling another adaptation, this one involving its enterprise resource planning (ERP) system.
Today Metcam uses its ERP system, Syteline by Infor, to schedule and track costs on the shop floor. But the manufacturing steps identified within the ERP still go by process-oriented definitions. To produce a part may require punching, bending, hardware insertion, welding, and painting, and to track orders, workers must log each job before and after each process. That’s a lot of data entry.
To simplify matters, the front-office team is attempting to change the ERP software’s process definitions to include a “cell” as one operation, each requiring a certain amount of time and number of people. This way, just one person needs to log jobs, and only when these jobs enter and exit the cell. This simplifies data entry and reduces the chance for errors.
The details get tricky because of the highly variable nature of job shop work. For instance, two adjacent cells on the floor have a punch press, several press brakes, and hardware insertion presses. Depending on capacity levels, workers may move jobs to other cells. For instance, say they punch and bend on one cell, but then find that the hardware machines are busy with a batch of complicated parts. So to keep parts flowing, they might send that work to hardware insertion machines in an adjacent cell. The company now is identifying all these potential routings. Ward admitted that it is a complicated task. But, ultimately, the system may help streamline operations even further.
For example, say one job requires the operator to contort the part awkwardly to insert the fasteners. Looking at the ERP data alone, a manager would see how long it takes for the hardware operator to make it through a batch of parts, identify that as a constraint process, and may add another hardware press to boost productivity.
Tracking time on a cellular level, though, can identify hidden opportunities. Consider the same part, but now the ERP system is tracking only how long it takes for the job to flow through an entire cell. The ERP identifies an inefficiency but just with the cell, not a specific fabrication process. The technicians at the cell see this inefficiency and brainstorm. What if a brake operator bent these pieces halfway through the sequence, went over to the hardware press to insert the fastener, then finished the bend sequence? It turns out this arrangement can increase throughput without additional people or machines.
This kind of thinking is what has made the cellular concept work so well, Ward said. The cells have been changed since the initial layout last year, and they undoubtedly will change again as new orders come in and the product mix changes.
Every change aims not to increase capacity of a certain cutting center, bending machine, or paint line, but to shorten overall manufacturing time. If more products don’t ship in less time, the fastest cutting, bending, or welding technologies really don’t make much of a difference.
“Since we changed to a cellular layout, we have never looked back,” Ward said. “My only regret is that we didn’t do this sooner.”
The FABRICATOR is North America's leading magazine for the metal forming and fabricating industry. The magazine delivers the news, technical articles, and case histories that enable fabricators to do their jobs more efficiently. The FABRICATOR has served the industry since 1971.