TAPPING HIDDEN CAPACITY: Increasing Throughput Via Changeover Time Reductions

TAPPING HIDDEN CAPACITY: Increasing Throughput Via Changeover Time Reductions


Greater choice means additional products and while some of the volume may be incremental, most of the time it also results in less of the baseline product. This means shorter production runs and a need to frequently change from one product or model to the next product or model. This results in a loss of capacity during the changeovers. Unfortunately, after some time, in most companies, it becomes normal and an accepted loss of uptime, which translates into a permanent loss of capacity. However, by employing changeover discipline and Single Minute Exchange of Die (SMED) techniques, the loss of productive capacity can be significantly reduced.

Another thing that routinely occurs is that changeovers become events separate from production. Sometimes it is an entirely different group of workers such as maintenance or a setup team that performs changeovers. These teams may have great experience in the mechanical aspects of a changeover, but not necessarily the operating parameters that enable the equipment to run smoothly. There is often a disconnect between both the line operators and the changeover team and then between the changeover team and the line operators. An example is that the production line has been operating successfully (high efficiency and low waste) for some time. The operating parameters are not recorded when the machinery is shut down. After the operators depart the line, the changeover team arrives (the equipment has cooled down) and begins the process of removing change parts for the last set up and installing the ones for the next production run. After the changeover team completes its tasks, the line is idle waiting for the next scheduled production run. This may be with operators from another production shift or, in some cases, from a different production line as the operators from the last time the production line in operation have been moved to a different production line in order to take advantage of both their experience and their remaining available hours within their scheduled shift.
Should changeovers be a separate event, it is highly likely that the start of the next production run will be characterized as inefficient and results in high waste, as the operators struggle to adjust the machinery to the ideal operating parameters. The loss of continuity by this approach results in further loss of capacity and results in greater costs.
Also, when changeovers are disconnected from the immediate production schedule, they become less time critical. They are conducted during scheduled downtime, which makes them more of an indirect task. As such time is allotted for the changeover, which means that there is little sense of urgency in completing the task and as in many things without that, work expands to fill the allotted time. This allotted time is now the expected time going forward, and there will be no effort to reduce the time; thus it is a built-in loss of capacity.
The appropriate metric for changeover time is good product to good product. Many times, as an independent task performed by teammates other than the line operators, it is described as steel to steel. That means from the time the mechanic touched the equipment to when the mechanic completed the task and moved on to the next thing in his schedule. Good product to good product means the time from the very last good product made as the line or machine was shut down to the very first good product manufactured. It captures the entire loss of capacity that the changeover created.

Early Career Changeover Frustration

My first experiences with changeovers as a production department manager was at a snack food manufacturing plant in west Texas. The plant had an unusual layout as many older manufacturing plants sometimes do. My production department was on one end of the plant, a distribution center in the middle, and the two other production departments on the other side of the distribution center. Another reason that the tortilla chip production department underperformed was that the maintenance department was located on the far side of the plant. So, a mechanic would have a significant walk to traverse the other production departments, the distribution center, and then the tortilla chip plant to investigate a downtime event. Oftentimes, he would arrive with the wrong tools or not have the requisite repair parts with him, and therefore, must make the trip several times all while the production line was shut down.

Due to the above limitation, we decided to deploy changeover parts in a storage box immediately adjacent to our production equipment.  In addition to the changeover parts, we stored routine repair parts that we might use on this equipment should a problem occur.  A spare set of tools was also placed in the storage box in case the mechanic would never arrive without all things being present.  Still, the time to transit the building was excessive and invariably, the mechanic was already working on some other issue in some other part of the plant.  To address this, we trained our operators to complete the changeover themselves.  To ensure that the changeover would be completed, and the equipment restarted by the operator performing the changeover, we moved changeovers away from shift change and towards the beginning of the production shift.  All other teammates were given specific assignments to assist the operator. Since it was our team performing the changeover rather than a mechanic, there was a sense of urgency and the changeover duration was reduced by well over 50%. More importantly, the startup was quicker with less waste.  The tapping of hidden capacity, in this case, was minor, probably around 2%.  The impact was primarily felt less over time.

Next Major Challenge

Moving north to Wisconsin with one of the snack food company’s best manufacturing plants was one of my major challenges. As a company, we had adopted a strategy of building fewer manufacturing plants and would meet our production needs during the summer months by extending our hours to 6 plus days per week, 24 hours per day (remaining hours used for weekly sanitation and maintenance of the equipment.). Under this new strategy, any downtime event would subtract from capacity and might result in shortages to our customers. In our potato chip department, changeovers from one style of chips to another were typically 45 minutes to one hour long. Again, the normal time to execute the changeover was near the end of one shirt so that we can reduce direct labor by having excess teammates clock out. Those that were doing minor tasks such as wiping down conveyors, sweeping the floor, purging the equipment of the other style of chips, and general clean up would inevitably perform those tasks in a lackadaisical manner as to just complete them. This lackadaisical effort quickly resulted in a reduced paycheck. Other tasks, such as removing the rotary slicing heads and setting up for the next product, were done sequentially even though there was an extra set of heads to set up for the next hour of production. This is due to the blades dulling and tearing the slices resulting in lower yields. Packaging machines needed to be stripped of the previous material, parts changed for the next sizes, and new material mounted and tested to ensure good package quality.
At a team meeting, I challenged the shift to reduce the next changeover from 45 minutes to 10 minutes and to perform the changeover mid-shift with maximum staffing (no breaks or lunches). We divided up the tasks and assigned specific members specific tasks. We doubled the number of teammates at the slicing equipment and agreed to track our time against a large clock that was visible in the department. Additional to the planning, I would reward the team (if we broke 10 minutes) with both pizza in the cafeteria and a gift certificate for dinner.
The next day, the team did not quite achieve the goal, as the changeover, measured as the last good product packaged to the first good product packaged, was completed in 13 minutes. Great, but not at goal. The team members had individual discussions with one another and asked for another opportunity. On the following day, the changeover was completed in 5 minutes and 45 seconds with great excitement by the team.

Changeover Reduction

Dramatic changeover reduction from 8 hours to 15 minutes in another operation, and from 4 hours to 1 hour in another and several other companies were accomplished.  Employing tools such as Single Minute Exchange of Dies (SMED) is very helpful.  However, they focus on only a portion of why changeovers have taken longer than necessary. These tools focus on removing all tasks that can be performed at another time being done outside the changeover window. These tasks include pre-positioning all change parts, tools, and other materials at the production line, installing quick-change connections, building a modular drop in assemblies, etc.  These are all useful, and I encourage their use.  What is missing is the team enthusiasm about a team goal and the reward system by management.  The other two principles worth repeating are (1) that changeovers are measured from the last good part to the first good part and (2) that there is an illusion of labor savings by reducing the team size.  In my experience, when you attempt to restart production, you need the entire team present and ready.  Waiting for the next shift or group of operators is often times inferior to the same team that shut down, performed the changeover, and started up the new product.  There is more of a sense of teamwork and urgency, and it is simply a smoother startup if done with the same team all the way through.

Hidden Capacity

Hidden capacity is all the existing time that is used for activities other than making the product. In many cases, it is a significant percentage of total time. Changeovers should be tracked as key metrics. “Things that Are Measured, Get Better” and “If You Cannot Measure It, You Cannot Manage It.”