Winning with Lean

Larry Martin hates the term lean manufacturing.

"When someone mentions lean manufacturing, it usually conjures up visions of layoffs and a stark, militarylike plant," said Larry Martin, vice president of Utah-based Martin Door Manufacturing Inc.

He blames lean manufacturing's bad reputation on companies that don't fully implement it. "Many companies see lean as a quick fix. They'll cut people, add quick die change, or implement some kanban principles and think they're through," he said. "In reality, it's a lengthy process, because there are factors known and unknown that are interweaving and contributing to your problem."

And he should know. His company's lean manufacturing journey began in 1991, and according to Martin, it's still just beginning.

Something's Gotta Give

Big players like Overhead Door, Clopay, and Wayne Dalton dominate the $10 billion U.S. garage door industry. Family-owned Martin Door Manufacturing, a manufacturer of high-end prepainted steel and specialty copper garage doors, focuses on the replacement door industry.

To compete, the manufacturer needed to carve out a market niche based on safety, quality, and on-time delivery. "Because our doors are the most expensive in the industry, people buy ours when they're fed up with the poor quality, unreliability of their current door, or have safety concerns," he said.

The company's steady growth and lack of strategic planning became critical in the early 1990s. Growing pains included production bottlenecks, high employee turnover, late shipments, high production waste, and quality complaints–all of which were impacting the bottom line.

"My cousin suggested looking into the quality movement," Martin said. "After doing some research, I found other companies had similar problems and there were ways to fix our manufacturing issues."

Martin's lean education started with a W. Edwards Deming seminar. "I learned education is key, and if you want to be effective in your plant, you have to obtain what he calls profound knowledge," Martin said. "Profound knowledge means that if you knew how to fix your problems, you would fix them. Deming recommends going outside the plant and bringing in thinking and knowledge you don't possess."

Because of lean's bad reputation, Martin involved all employees in the education process. "For this to work, we needed employee buy-in. We held seminars to educate everyone on lean principles and how we can improve the manufacturing process," he said.

Once the team was educated, Martin began systematically to heal the company's growing pains.

Meandering Production Flow

Over the years the plant grew from 30,000 sq. ft. to 250,000 sq. ft. The layout was a maze of connecting rooms, with machinery and assembly processes scattered throughout the plant.

A consultant recommended Martin start by measuring the plant's value stream (the point at which raw material enters the plant to where it leaves as a completed product). "The string was thousands of yards, and production didn't flow in a straight line," Martin said.

To streamline production flow, all mechanical presses were moved to one centralized area of the plant, as well as raw materials. Three major assembly areas of windows, door panels, and hardware supplies also were consolidated into areas with their raw materials.

"We tried to create the shortest path between the back door and the front door, which resulted in cutting production flow by two-thirds," Martin said.

Overhauling Assembly Operations

Every completed door is shipped in two packages: the panels and the hardware (tracks, springs, rollers, and hinges). "Each order is unique, so we're like a job shop," Martin said. "We learned early on to plan and think not like we're making one of a kind, but producing a commodity–everything needs to be flowing and moving, and you shouldn't have to concentrate on any one piece."

Before implementing lean manufacturing, the company approached its hardware boxes as an entity unto itself. When components were running low, a worker wrote up an order and sent it to materials storage. Someone would get on a forklift, retrieve the materials, and maybe get it there on time. Orders usually were late, he commented.

To maintain some sort of flow, Martin bought a Kenway stacker (automated storage and retrieval system). "I tried to have a month's supply of everything on hand to avoid shutting down the assembly line," Martin said. "I thought I had a cool thing, but a consultant told me I had inadvertently created lots of unnecessary inventory."

Martin's consultant recommended using the kanban system, a Japanese concept from Shigeo Shingo that means empty bin or move card. The system ensures an uninterrupted supply of parts without complicated ordering procedures. The number of parts in a bin is based on the highest expected usage rate and the longest resupply time. Bin sizes are determined by bin quantity and part size.

The company now uses a visual pull, two-bin kanban system in a straight line and a streamlined U-shaped assembly line for packaging small parts, which operates behind the assembly line. As bins are emptied, workers pull forward another bin with the same components, which leaves an empty space in the back. Empty bins are placed on a slide behind each workstation.

The empty spot is a visual reminder to replenish the assembly line's supply. The card on each bin states what should be in it, and twice a day people from storage look for empty spots and replace Gaylord bins of bolts, for example.

"No one is on the floor writing down they need parts because the kanban automatically flows and keeps it filled," Martin said. "Instead of having a 30-day supply of inventory, I now have a 30-minute supply on the line."

Die Change Nightmare

In the 1980s raised panel doors became popular, and the manufacturer wanted to minimize its new tooling investment for a 500-ton mechanical press. "We had several different raised panels we needed to stamp. To save money, a toolmaker suggested building one die with movable components within the die. So, when you want to make a die change, you pull the die out with a jib crane, open it, pull out pieces, and put them back in new locations," Martin said. "It turned out to be a nightmare because it took three to four hours to change die components."

Because of cumbersome die changes and hours of downtime, the company ran each die setup for eight hours to two days, which created an abundance of inventory. "For one die setup, we created two weeks' or more of inventory," Martin said. "I didn't know this would be the end result of my nifty die.

"Inefficient die changes impacted every aspect of the shop floor, especially shipment deadlines." We would promise a dealer a date, but because of tedious die setups and lack of commonality between the different models, we couldn't get it out on time," Martin said. It took up to a week for normal orders to be manufactured and shipped, he added. Special orders took even longer–two to three weeks for completion.

Lengthy die setups also contributed to excessive waste. "When you have so many die setups that are uncontrolled, the first part out is never good," Martin said. "By controlling our setups with a set value instead of an artistic-type adjustment by the operator, our waste went from 4 percent to 1.5 percent. Now our first part is always good."

To reduce downtime and inventory, Martin incorporated Shigeo Shingo's single-minute exchange of dies (SMED) principles. SMED analyzes the changeover process by encouraging shop personnel to determine why the changeover is taking so long, and how it can be reduced. As a result, die changes were reduced from three hours to three minutes. "To achieve SMED, we invested in five new dies. The money we saved in inventory paid for the dies," he said.

Consistent with SMED principles, Martin assembled teams of operators and plant managers and gave them a specific area to study and improve. "I challenged the press team to see if they could improve upon the three-minute die change for our smaller 200-ton presses," Martin said.

Four months later the team reported they beat the engineers and created the shortest possible setup time–zero. Martin was skeptical but willing to see what they had done.

"The team took four or five different dies and permanently mounted them on the large bed of an idle 200-ton press at the right shut height in fixed positions. Whenever they needed to make a part, they could switch out dies instantly," Martin said. "It was genius."

Because of SMED, the company measures defective parts per million instead of per hundred. "Now we rarely ship late, I have no back orders, and the product is right–the first time," Martin said.

Conquering High Employee Turnover

Prelean Martin Door Manufacturing had an annual employee turnover rate reaching 50 percent. "Again, lack of strategic planning was hurting us," Martin said. "We had no published salary range, raise schedule, or company review policy."

To show employees what they could expect in the future, the company assigned every job a pay range, with a median based on wages in similar industries in the Western states. A review system was established, and now all employees meet with their supervisor every three months to discuss future goals. Raises are automatic, and additional raises are based on acquiring new jobs skills. The company instituted cross training and an annual bonus based not on merit or pay but on equality. "If we make a profit, a certain percentage of sales is divided equally among all employees," Martin said.

Over time all the changes raised employees' satisfaction levels and lowered turnover to below 20 percent per year. The manufacturer now has numerous employees who have been with the company 15, 20, and 25 years.

Lean on Steroids

During the shop floor's metamorphosis, Martin found the term lean manufacturing didn't fit with what he was trying to accomplish.

"I didn't want the negativity [layoffs and a stark office] that term carries," Martin said. "We were incorporating teachings from several lean gurus, and our goal was to keep the plant flowing, so we coined our own term–Flo-Max™–lean on steroids," Martin said.

To document the company's progress, a one-year rolling sales forecast was established, in which sales revenue and units produced for each model were projected. "From that forecast we perform capacity planning, as well as project profitability and inventory levels, machine use, and daily production levels," Martin said.

Supervisors post units or parts they're scheduled to make that day and report if they did or didn't make goal. "If a production goal wasn't achieved, was it because someone is out, a machine is down, or materials were late?" Martin said. "We track everything. If there are any production problems, we use an appraisal chart to address the worst one first, and after awhile they're gone."

In February 2004, 13 years after adopting a lean philosophy, Martin's Flo-Max strategy helped the company reach a key milestone. "We became the only manufacturer in the world, to my knowledge, that can take an order for any metal garage door and get it out the door in 48 hours," Martin said. "This is possible only because we had vision and allowed the improvement process to continue until we achieved it."

Satisfied with the progress on the shop floor, Martin is focusing on Flo-Max for the company's office. "I asked the office staff to examine every piece of paper they produce and evaluate all meetings and ask why they are doing it," Martin said.

Meeting times now are dramatically shorter, and the time savings eliminated the equivalent of a full-time employee. The company also reduced its consumption of paper by 400 reams, and according to Martin, the process has just begun.

Martin is proud of his company's lean strides but cautions stampers to take it slow before making radical changes.

"If you don't approach lean with careful planning, education, and an unwavering commitment, you won't make it," Martin said.

Martin Door Manufacturing Inc., P.O. Box 27437, Salt Lake City, UT 84127, 801-973-9310, fax 801-972-5809, www.martindoor.com