The Columbus McKinnon of Mexico story

Columbus McKinnon of Mexico, originally known in the marketplace as Endor SA de CV, has always been a well known name in the lifting equipment business in Mexico. A number of years ago, Columbus McKinnon of Amherst, N.Y., purchased part of Endor, which at that time had been a family-owned and operated business. 

More recently Columbus McKinnon purchased the remaining portion of the business. In June 2000, Michael Garmon, formerly an internal auditor at CM was appointed as general manager of Endor SA de CV. Shortly thereafter the company changed its name to Columbus McKinnon of Mexico.

Endor had just gone through a rocky systems implementation. Operations, inventory and administration at the plant were spread between six buildings. The warehouse was full to the ceiling and wip clogged the floor. On time shipments were typically in the 40 percent area and MRP and inventory accuracies were in ashes.

Two years later, inventories fell, on-times ranged in the high 80s to low 90s, MRP and inventory accuracies stayed under control. 

Simpler Consulting's John Kim, Columbus McKinnon hoist division group leader Dick Davidson, and Columbus McKinnon CEO Tim Tevins paid a visit to Endor to kick off the Lean program at Columbus McKinnon Mexico. 

At the end of June 2002, Simpler Consulting senior consultant Bill Supple went there for an initial value stream analysis. Garmon set forth his six- to 12-month objectives for Columbus McKinnon of Mexico’s lean program.
• All hoist assembly, machining and inventory under one roof.
• All chain operations under one roof.
• Inventory reductions of 10 million pesos.

The first two objectives have already been fulfilled, with the third objective well underway.

Garmon appointed Jonathan Splittgerber to take leadership and responsibility for the Lean program at Endor and later appointed Alejandra Gonzalez Bustamante as a Lean coordinator. Jonathan worked at Endor since February on assignment from Columbus McKinnon to help make the ongoing system implementation successful. 

Garmon appointed Splittgerber as materials manager when he was appointed at general manager. Bustamante was the secretary for the previous general manager and stayed on with Garmon  in the same capacity. The Lean process in Mexico proved that bilingual coordinators are needed, which made Splittgerber and Bustamante ideal choices.

Columbus McKinnon Mexico manufactures chain as well as manual and electric hoists. When it started Lean, the chain plant was a small part of the business. The bang for the buck was in the hoist plant.

The first event was set for Aug. 15, 2002. Simper’s Lean documentation and team leader guide was translated into Spanish. CM set up a team to work on cycle-time reduction in the forge, to eradicate shortages which affecting the assembly department.

Half of the workers produced twice the output and the appearance of the area radically changed for the better.

“I knew that we could make it better, but I didn’t know it could be made that much better in so little time for so little money,” said Garmon. Since this first event, Garmon has supported the Lean program 100 percent.

CM had a vision, which became better defined over time, but the direction was unmistakable. The vision at the time was to make machined pieces flow better into assembly. The next stage in the plan was to run and better understand set-up reduction events.

Supple advised the workers to shift focus to assembly because it would expose the hidden problems in the machining department. CM Mexico assembles electric hoists and cable pullers from pieces machined in its own shop or machined at diverse CM manufacturing locations. 

Bill’s guidance sent workers into the assembly department for a one piece flow event for the 640 puller, which is machined at Virginia hoist and assembled at CM Mexico for the national market.

The assembly team achieved one piece flow ,a productivity gain of 269 percent, and a reduction of distance walked by the operator of 99 percent. The forge team, focused on setup reduction, reduced the setup time for the forge hammer from 127 minutes to 15 minutes. The team also faced the problem of preheating the die which took 45 minutes. The team implemented a solution which allowed the die to be preheated in one minute. The distance the operators walked in completing the setup was reduced by 88 percent.

For the next events, the lean team appointed three teams. They chose two shop floor events and ran a third team on an administrative topic. For a small plant like Columbus McKinnon of Mexico, two events are very taxing for the tool crib and maintenance areas; in this way we could maximize the use of the consultant’s time without overtaxing the support departments.

Building on the success of the assembly event, we began a series of one piece flow events for all three of Endor’s cable puller models. For these items, most of the pieces are produced in the machine shop, running an event in assembly exposed many of the problems that were hidden in the machine shop. 

The team learned that much of its equipment is old and can’t hold the necessary tolerances. The team began by checking all pieces to specifications and the pieces out of specification were systematically corrected in the machine shop. Assembly time was reduced from 53 minutes to 17 minutes, a gain in productivity of 211 percent and time to pull the work orders of three hours was eliminated with virtually all the work in process.

Columbus McKinnon of Mexico also produced pieces for its sister company Abell Howe. Abell previously outsourced these pieces from a supplier in the U.S. CM of Mexico outsourced cuts and bending of some of the heavy steel. 

Chief engineer Victor de la Fuente and Splittgerber previously completed a cost feasibility study on bringing these operations inside the plant. At the time of the event, the new equipment was already on the floor and the processes were all new. As the event progressed, it became very obvious that the scope of the event was too large, so the team focused on doing one sub-assembly well and a follow-up event was scheduled.

The team achieved a 316 percent productivity gain on this sub-assembly as well a 68 percent reduction in floor space utilized, as well as a 97 percent reduction in distance walked by the operator. A nice compact manufacturing cell for Abell Howe products began to emerge.

The third team focused on reducing the time involved in processing accounts payable. This team analyzed current procedures and policies and tracked supplier invoices through the system. By the end of the event, they revised and implemented improved procedures and policies resulting in reduction of internal waiting time of 97.3 percent and the actual payment processing time by 72 percent. What had been a constant, week-long process was reduced to an activity which was started and finished the same day. 

The vision began to develop of an assembly area with a gravity feed parts system that would react daily according to pull. The next team continued with the next model of cable puller, this time achieving a 94 percent productivity increase and the distance the operator walked was reduced by 98 percent. Floor space was reduced by 52 percent. During this event, the team achieved the current layout for all three models.

The Abell Howe team continued with the one-piece flow event. It achieved one-piece flow with a 378 percent increase in productivity, achieved another 43.8 percent reduction in floor space and reduced the distance traveled by the operator. Previously, when orders from Abell Howe were processed, there was a chain of bottlenecks choking the plant. Now after the two events, even though not every Abell Howe package has been leaned yet, what we learned from these two events have made Abell Howe into a reasonable blip in our production instead of a massive disruption. 

The third team focused directly on the process of the accounting month-end-close. The goal was given to the team to be able to close the books in four days. The tasks required for month-end-close were evaluated and the distribution of the workload in the department was evaluated. The team redistributed workloads, and started development on a mechanism to get information from CM Mexico’s payroll system in a more accounting friendly manner, reducing the processing time from one day to two hours. Overall, the team reduced the month-end-close process from seven days to four and a half days.

We appointed an administrative event to reduce the amount of time in preparation of shipping activities. The team analyzed the waste and factory flow, and recommended that shipping be moved. Garmon gave the team a thumbs up, and the shipping department was left empty and spotless by the end of the week. All inventory in the shipping department was absorbed into the assembly line. The team also relocated the invoicing function in the shipping area. The new location for shipments is much closer to the assembly area and a 65 percent reduction in floor space was achieved. The new layout meant fewer forklift movements and that shipping was more manageable with fewer people.

We tackled the last cable puller model with a 64 percent increase in productivity and distance walked reduced by 64 percent. This product line had every model addressed. The assembly team left a work cell that can do a bench changeover in fewer than five minutes. The sales department commented that the deliveries and on-times of this product line became more robust and reliable.

For the last internal event of the year, we ran an event to relocate the tool crib and maintenance and a Six Sigma event in assembly. The Six Sigma in assembly, plus relocation of the tool crib and maintenance, which were needlessly taking up space in this area, freed up space in assembly. The Six Sigma team accomplished a lot, but we realized how much more was left to accomplish. 

A follow-up Six Sigma event was scheduled. In the follow-up event, the team determined that the best way to accomplish the Six Sigma objectives was to move the rest of the warehouse. The follow up team moved 13 racks, more than 20 tons of material, relocated two assembly areas and freed up 3,300 square feet of floor space. The desk areas for the warehouse people were also relocated to the assembly floor. All the inventory moved to the assembly area was left neat and orderly by the team and in a viable layout for follow-up with one-piece flow events for the rest of our assembled products.

At this point we can see that as we progress in the Lean process, all assembly, machining and inventory related to the hoist plant will fit into what was originally just our assembly area. Our warehouse now stands completely empty, ready for a more productive purpose than housing inventory. We can picture gravity-fed assembly benches with the ability to switch between items manufactured to handle the low volumes and high varieties that our customer’s demand. We can picture our departmentalized machinery dissolving into a few critical cells and being linked to assembly by pull systems. We can envision set-up reduction events continually decreasing the need for racks and pieces in assembly. 

We can see that once we have done more events in assembly that the response from assembly will be so agile that we will be able to eliminate finished product in shipping. We can begin to see the disappearance between the traditional departments such as purchasing, production control, shipping, receiving and stores. Some of the benefits we have already started to enjoy as a by-product of our lean efforts. 

The chain story
As much ground as we feel we are gaining in the hoist plant, the Lean core team found it necessary to switch focus. We received the news of the Cobourg, Ontario, plant closing and the plans to move all grade 30 and 43 volume here to Mexico. 

We were going to receive 10 more chainmakers. We were to be ready in mid-February 2003 and the chainmakers would start arriving at two week intervals. We were not ready. 

We did not have enough trained operators, a situation aggravated by high employee turnover. Experience had taught us that trucking the chainmakers across the continent could render them impossible to set up and run quickly, plus we would have insufficient manpower in the tool room and maintenance. 

Extensive installations would have to be completed so that pneumatics, water circulation, electrical and floor space could be ready. Tooling and planning issues would have to be resolved. The floors themselves were not reinforced concrete and enormous caverns had developed under them over time. The river behind the plant was threatening to topple the back wall of the plant and flood the chainmakers. We knew that we were going to have to limit our focus almost exclusively to the chain manufacturing area in order to survive the transition. It was evident to us that the process was out of control and the mechanical integrity of the machines was questionable. 

We were already in trouble because we couldn’t keep the machines running and couldn’t meet deliveries because of it; now our productive capacity and demand was to double in a short space of time. 

We began to think about what we know at Columbus McKinnon Mexico about making chain. Senior consultant Bill Supple challenged us to think about our base assumptions about chain. He asked one simple question, “what do you make here?”

Everyone insisted that we make chain. Bill insisted that we make links, one perfect link produced over and over again is a perfect chain. Master the link and you master the chain.

We had to go back and think about the perfect link of chain before we could go on with chainmaking. We reviewed our specifications for raw material, our draw block, our in-process dimensions for formed chain, and our in-process dimensions for welded chain. The lack of information was dismaying. 

Supple asked us, “Why do you do the things the way you do?”

“Because that’s the way our sister company does it,” we answered. 

“Errrrr!” said Supple. “Not very smart.”

In the chainmaking process, what we thought we knew is that the operator just has to know. He learns this from spending a long time with the machine. He knows whether his formed chain is right by feel; he knows that his weld is neither too hot nor too cold by the color of the glow.

In our predicament, it was obvious that we did not have the time or knowledge to develop operators this way. Our turnover had already taken an incredible toll on our pool of skilled operators. We had to figure out a way to quantify the process. We had to find a way to prove whether the machine was running right or wrong. We had to find a way to put the machines in a state of mechanical grace so they would repeat and eliminate all the wild variables resulting from the sorry mechanical condition of the equipment. We had to find a way to train our operators, not in one or two years, but in three weeks.

We received news of two machines sent from the Canada plant to our location to run Grade 43 chain. A complete revision and rebalancing of substation transformers had to be done. Loads were balanced and previous high voltage installation mistakes were corrected. 

Chief engineer Victor de la Fuente had to complete the projects on Sundays and holidays. Since it was the rainy season, his high voltage crews had to work underneath tarps to safeguard against the risk of a sudden rainfall during a sensitive high voltage installation.

In June 2002, the machines arrived from Canada. The machines were placed in position. Both machines had severe problems, mostly electrical and setup problems. On one machine, the welding transformer had a major short; on the other machine we ended up rewiring the entire wiring harness on the welder.

The machines were finally running intermittently after a period of four or five weeks of maintenance working double shifts, and took even longer to settle down and produce any reliable production. Our maintenance crew was as ready as they could be and we had an operator/supervisor from Canada with two decades of experience on the machines.

Despite our best efforts and preparation we could not be ready on time. We fell behind on deliveries of the new products. This taught us an important lesson about what was to come.

We took advantage of a holiday and ran a total preventative maintenance event on one of our chainmakers. The team disassembled and completely rebuilt the former and one of the two welding heads of our 3/8-inch chainmaker. Detailed notes were taken.

Pieces which could not be replaced immediately were given the best interim solution. The pieces that we had to patch up were machined in the tool crib for the follow-up event scheduled. On the follow-up event, the new pieces were installed and the second welding head was rebuilt. The team took measurements of chain off the former and found no more than .001 to .002 thousandths of an inch in the dimensions of the link.

Shortly thereafter, we received word that the Canadian plant was closing and that we would be running an additional 10 machines. On Jan. 3, we received approval from corporate for the project.

Our deadline to install chainmakers was Feb.14. We would need to provide for larger wire storage, provide space for chainmakers and support equipment. We would need to correct all past errors including the floor and the back wall of the plant that the river threatened to topple. We needed to complete water circulation, pneumatic, electrical, structural and cement work for the floors and shoreing up the wall on the side of the river.

In order to pour the floor, all the machines would have to be moved. Garmon was very clear, “No machine goes down any longer than necessary, for no reason can we shut down the entire plant”.

Faced with extremely short deadlines and an enormous project, Fuente got a team together and developed a solution for moving machinery rapidly without causing a significant interruption.

A chainmaker consists of a former, a welder and a control panel for the two. In order to move a chainmaker ordinarily the three piece are disconnected moved individually and reconnected. The distance between the former and welder is also an important factor in machine setup. There have also been cases where even short distance movements have taken three or four weeks to put the machinery back into production. Victor’s team developed a sort of platform that allowed the former, welder and control panel could be moved at once. Construction proceeded under the direction of Fuente.

A concrete floor was poured for the new area that was to become wire draw. A tent was rented and place in the new area which did not yet have a roof and machines were moved one by one each machine going back into production in its temporary location before moving the next machine. In five days, all 10 chainmakers were moved and put back into production. 

Not so much as one machine had a major failure due to the movement. All the machines continued producing and even lean events on the machines were run in the midst of the movements and in their temporary locations. The installation of the floors crisscrossed the electrical and pneumatic and water circulation in the ceiling above. Crews worked simultaneously across the floors and ceilings even as new operators were being trained. The second section of floor was poured. The machines were moved back, again in five days and without major problems. 

On Feb. 14,  the construction was proceeding according to plan. We were ready for the machinery, even though the construction was proceeding. Extra material was drawn and stored so that the draw block could be moved; if the drawblock went down it would stop the entire plant. The drawblock moved without incident and the third section was finished. 

Garmon, Splittgerber and Fuente traveled to the Lexington, Tenn., and Cobourg, Ontario, plants to evaluate the equipment coming down. At the time they arrived in Lexington, they wrapped up one of their own Lean events with Supple.

In Lexington’s event, they had proven that if a chainmaker is set up properly and left alone it will run very well with significantly higher productivity. This may seem simplistic, but a culture had developed and a psychological need had developed on the part of the operators that they weren’t doing their jobs if they weren’t continually futzing with the machines.

Higher productivity was a result of obtaining a good setup and just leaving the machine alone. Lexington’s event had doubled production on the machine. Garmon was enthused. Immediately on our return to Mexico, he directed that a team be put together to duplicate the results achieved at Lexington.

Two teams worked side by side and ran a similar event as the one in Lexington. Five Machines were painstakingly set up. The setups were documented and a new setup format was developed using photographs posted at the machine to make it as explicit as possible for the new operators. 

The machines were tracked and our teams proved the same results as the original team at Lexington.

Through the use of production control boards at every machine, we began to develop a body of information about why our machines were down. At this point, we were focusing more on the perishable tooling which must be changed on a continuous basis. We reasoned that if we "Leaned" tooling changeover processes, we could reduce total downtime and better assure that the tool changes were performed correctly.

The team categorized downtime by concept and attempted to slay the biggest dragon -- electrode changes. As the electrodes weld, the links wear an impression into the face of the electrodes. On each head, all four electrodes have to be adjusted precisely and the tensioners must be adjusted properly.

In our first method, we removed the electrodes, refaced them and cut grooves in them with a hacksaw to keep the links from sliding against the electrodes. Sometimes the electrodes had to be reworked three times to get them to function right. It was evident that better trained operators might not have so many difficulties, but we could not rely on difficult-to-learn artesian methods.

The team figured out a way to position the electrodes precisely each time by adding a stop in the back in the electrode holders. A screw was added to the back of the electrodes as a spacer so that despite electrode wear and refacing, the electrodes going back in the machine would always be the same length. A cutter was hand-machined to cut precision grooves in the face of the electrode. With the efforts of the team, the electrodes can be changed in four minutes instead of 45 minutes depending on the experience of the operator. 

We are still testing different machining patterns to find the optimal pattern. When we stop to think that every shift is changing electrodes, when we have this implemented across all machines, this event alone may increase total uptime as much as 10 percent to 15 percent.

In the last event, we put together teams to organize the flow of work on the floor and to develop greater controls for scarp. Another team developed a pit-stop program so that tooling changeovers could be put together faster. A third team worked on one-piece flow of the grade 43 binder assemblies achieving a 200 percent gain in productivity. The team wrap-ups were very positive. 

“I see a turning point,” said Supple. “The teams and the people on the floor are now telling you what they need to do their job better. Don’t let them down”.

The lean events in the chain plant are just starting to pick up steam and generate inertia. We already saw significant improvements in productivity. The confidence of our Lean teams has increased greatly. We have Lean converts on the floor and the direction that people are going is starting to take shape. Many gains in productivity were made and many more were planned.

Back to top

Back to Web-exclusive articles archives