6 The FABRICATOR Industry Research Project NOVEMBER 2017
AS OF SEPTEMBER, C.O. W.
HAD ACHIEVED NEAR-PERFECT
ON-TIME DELIVERY FOR THE
YEAR—AT 99. 9 PERCENT.
THAT’S A FAR CRY FROM THE
1980S, WHEN THE COMPANY’S
WAS AN ABYSMAL 65 TO 75
How to achieve on-time-delivery perfection
Vice President of Manufacturing
C.O. W. Industries
Gary Carter has been to his fair share of classes on continuous improvement: lean manu-
facturing, theory of constraints, Six Sigma, and more. “Nothing is perfect,” he said. “We just
take what fits best from each.”
Nothing is a clean fit because no two fabricators are exactly alike. The company moniker,
C.O.W., stands for Central Ohio Welding, which comes from the original welding distribu-
torship the founders started more than a century ago. Today the company has laser cutting,
punching, bending, and all the other common fabrication processes, plus a healthy dose of
For years shop floor workers couldn’t see much of, well, the actual floor. Metal blanks and
half-completed parts lay everywhere. People looked for parts constantly, and all too often
ended up remaking them. This was reality for the shop in the 1980s, when the company’s
on-time-delivery rate was an abysmal 65 to 75 percent.
The reality is much different today. The entire floor is impressively devoid of work-in-process—except in front of the welding operation. The WIP before welding shows how the
shop has applied the drum-buffer-rope technique of the theory of constraints. The “drum” is
the constraint process that dictates the throughput of the entire shop. For Carter’s operation,
this is welding.
The “buffer” is a defined amount of WIP that feeds into the constraint, as outlined by tape on the floor—what Carter calls the “
parking lot.” The state of this parking lot provides a signal (the “rope”)
for Carter and others to release orders. When the parking lot is full,
they release only those orders that do not require welding. For the
rest of the operation, Carter and his team manage the flow carefully
so that jobs move quickly from one manufacturing step to the next.
Making this possible is the fabricator’s repeat orders, scheduled
based on variations of a kanban replenishment system. The specific
arrangement depends on the customer, but the basic idea behind all
of them is simple: a signal (newly delivered container from the customer, for instance) triggers C.O.W. to replenish certain products. A
large amount of repeat orders also makes setups more predictable,
repeatable, and efficient.
During slower times, Carter makes sure to build a little extra inventory for certain customers. He has the blanket order, so the shop
isn’t assuming the risk of obsolete finished goods. This level-loads the schedule and frees
capacity during busier times. Carter uses this capacity to serve what he calls “job shop customers,” those nonrepeat orders that come in throughout the year.
“This overcomes the feast and famine nature of this business,” Carter said. “When cus-
tomers send in everything at once, no one is happy, because you’re not meeting the shorter
lead times that everybody wants now.”
Carter conceded that the lean journey hasn’t all been smooth sailing. For instance, for a
time the fabricator ran another buffer in front of the press brakes. “That basically allowed us
to run too much,” he said.
Mishaps aside, the system is evidently working. As of September, C.O.W. had achieved
near-perfect on-time delivery for the year—at 99. 9 percent.