A story about setup

When North American executives started going to Japan to tour plants, having heard of the “Japanese miracle”, one of the things they wanted to see was quick die change. Many did not believe that the Japanese were telling the truth about press setups in less than ten minutes, when their own best efforts were not yielding anything like it, and their own people were telling them that it was impossible.

One group of executives invited to see such a die change, and to be present at a specific stamping press in the plant they were visiting 11 am, when the next die change would take place. Since they were not paying very close attention to the time, they got to the press about 15 minutes late, and all they saw that day was some cleanup activities around the press, which by that time was already making parts. Astonished, they asked what had happened. They were told that, as scheduled, the press was changed over precisely at 11 am. They believed that 11 am was just a target, not a fixed time in the daily plan. They were informed that, had the setup crew waited for the executives to arrive, the rest of the day would have been thrown out of whack for a large number of people who depended for their work on parts flowing from the press in precise quantities, at precise intervals.

The lesson they learned was that quick changeover exists in an environment of precision and standardization, that it is not an isolated aspect of the production system.

Role of skill

A key issue that will be confronted when reducing setup is the role of skill – and those that possess skill. The skill in question is the ability to get a piece of equipment or system to function properly after it has been changed over. As was discussed, many setups suffer from variability in both the tooling and the machine. Those with skill in setting up have the ability to quickly get settings correct, under the specific operating conditions of tooling and equipment. Material also suffers from variation, as does the environment, and all of these factors must be put into harmony by the skilled setup person or team.

SMED strives for a setup that requires no adjustment, no measuring, indeed, for a simple procedure that has been completely standardized. The skill that was formerly required to set up will no longer be needed to achieve a successful changeover. To speak to the concern that we normally feel when we take away the skill and replace it with standardization, it should be said that that skill must be “vested” in the setup – the knowledge of the factors of variation must be made explicit, and eliminated from the setup. How those who have performed the setups are involved in reducing the changeover time can be key to how well this reduction in achieved. Of course, engineers and others can study the elements of the setup, and determine how to control them, but such studies tend to be lengthy, and the results slow in coming.

What about afterwards? It has been said that no one will help to improve a process, if the result is the elimination of their job. First of all, the key result of setup reduction is not elimination of labor. Labor is simply just not that big a part of the total cost of production that it should be pursued as a primary goal. Of course, the history of industrial societies is one of less and less labor required per unit of output. So, labor reduction is inevitable. What needs to be understood, however, is that some of the labor saved needs to be used elsewhere to maintain, and improve, the new system created through improvement processes such as SMED. It is prudent, then, at the outset of a SMED program to assure those involved in changeover, be it operators or dedicated setup resources, that there will be no layoffs or terminations as a direct result of the program. Normal turnover, training programs aimed at those who have the ability to gain new skills, and other ways of redeploying people, are the best way to allay fears about involvement in SMED.

SMED and TPM

Standardization is not just for tooling or the element that is being exchanged. It applies to what is not being exchanged, the machine or permanent element. Adjustment becomes necessary if the exchanged element, or the permanent element, or both, have variation in performance. It is especially difficult to get right if both are varying from one run to the next.

Standardizing the tooling or exchanged element will not eliminate adjustment if the permanent element, a machine, system, or program, is not stable and standardized. There is thus a link between SMED and TPM, between setup reduction and maintenance. Wear, vibration, and other kinds of drift from the design condition all create problems that must be dealt with during setup. This kind of instability is particularly troublesome, because they cause the internal aspect of the changeover to take more time. If the condition is not known, or not dealt with during scheduled maintenance, then it must be dealt with, whether directly, or provisionally, during changeover. The second approach is, unfortunately, more common, because of the rush to get back up and running. Thus, the problem is faced again and again, with no chance for standardizing the internal setup procedure.

Careful observation is required to disentangle the sources of variation. Preventive maintenance and equipment improvement are necessary to eliminate variation in the permanent element of the production system.

Elimination of Adjustment

Adjustment is a major source of downtime due to changeover. It is closely associated with conversion of internal to external – the adjustment is done externally. One of the basic ideas behind conversion from internal to external is elimination of assembly during downtime. By a simple exchange of preassembled components, setup is reduced.

To eliminate adjustment, the preassemble component is adjusted externally, or all assemblies are standardized, so that no adjustment is required. Examples of external adjustment are pre-heating of plastic injection molds, and shimming of dies in a trial press. Examples of standardization to eliminate adjustment is using a common die height for all tooling going into a specific stamping press or injection molding machine, and the use of keyways and locating pins to place tooling in a precise position without the need to measure off a datum point.

In most cases, it will require detailed study to determine the relationship between final settings and the conditions of the setup – what factors are influential, and how to measure and adjust for them prior to starting the setup. Keen observation, statistical analysis, and trial and error are all useful approaches. The key approach is to insist that it is possible to understand the relationship, and to make the necessary adjustments beforehand. Chemical processes can be particularly difficult to understand completely. In many cases, simple approaches, such as cleaning, and maintaining a constant environment, including temperature and humidity, will go a long way toward reducing and eliminating the amount of adjustment required as an internal activity.

Separate internal from external - not as easy as it sounds

We have indicated that the simplest and most effective way to reduce setup is to separate internal from external. But we have also indicated that there are many reasons why this does not happen. Some of the key reasons are:

1. The person responsible for the setup is also the operator/value creator. This means that there is no time to carry out the external tasks while the operations are in process, since the operator is engaged in value creation. Many organizations are not convinved that a separate setup function can be justified, especially if customer demands are being met. Some organizations designate setup an “indirect” function, and follow a management metric regarding the ratio of “direct” to “indirect” labor – the higher the ratio the better. Another reason given for operators to perform their own setup is that this places full responsibility for the setup on one person. When efficiency or performance to standard is used as a measure of performance, it is thought that setup personnel will rush to complete the setup, and may not pay sufficient attention to quality.
2. The scheduling system has a “standard” for setup, and the standard is being met. Consequently, there is no perception that setup needs to be reduced, especially if large lots are not understood to be more costly that small lots. Furthermore, the perception is often that faster setups are synonomous with simply working harder, or rushing and cutting corners. If customer demands are being met, the standard is unlikely to be reviewed. The fact that the standard may have been set without significant study, or even arbitrarily, is usually quickly forgotten by those setting up the scheduling system, and completely unknown to those who come after. It is even the case that when the standard is not routinely met, the extra time taken to set up is allocated to some other aspect of the reporting system, usually breakdown.
3. For various reasons, what to do next is not communicated to those responsible for the changeover. Often, expediting is the way the next job is selected. This can be a result of “first come, first served” – reasonably so in, for example, a restaurant. Or perhaps there are constant change to priorities – again, reasonably so in, for example, an emergency ward of a hospital. On the other hand, in some instances, each sales person or manager believes that his or her job is more important, and selection is a result of last minute negotiation. It may also be the case that jobs take an unknown amount of time to complete, perhaps because of material shortages, breakdowns, labor shortages, and so forth. In this instance, the perception is that scheduling is futile, and one might as well wait until the job is complete before deciding what to do next.

It may thus be the case that there is a lot of change required, before the simple rule of separation of internal from external can be applied, to reduce setup. A strong vision of how to get to the point where setup reduction can be applied to significantly improve customer satisfaction while improving profits is required. This vision is encompassed by the term “Lean” or “the Toyota Production System”. We thus refer to “lean healthcare” or “lean manufacturing”, and other lean business systems. The key element is “flow”, or lead time reduction.

Steps to shorter setups

We have seen that separating the internal and external elements of a setup, and carrying out the external elements while the value is still being created, is the first and usually largest part of reducing setup time.

The complete list of steps to minimizing setup time is:

1. Separate internal from external, and carry out external operations while value is still being created.
2. Carry out the internal operations as several simultaneous operations.
3. Convert internal operations to external operations.
4. Simplify and automate all remaining internal operations to speed them up.

Briefly, these steps are understood as follows.

Carry out internal operations as several simultaneous operations: This is often referred to as “parallel work”. The typical image is of the race car pit stop, where four tires are changed, fuel is topped up, the wing is adjusted, and so on, in just a few seconds. Anyone who has watched a car race will have seen how this is done – utilizing five, six, or more pit crew. By analyzing the tasks of the pit stop, and dividing the work into tasks that can be done simultaneously, the internal work is done in a fraction of the time it would take if only one task was done at a time.

Convert internal operations to external operations: Many internal tasks consist of replacing and assembling several parts of a machine, a file of information, a kit of some kind, and any other setup that has several components. Take the above example of the pit stop – a tire change consists of replacing a tire mounted on a rim. By mounting the tire on a rim prior to the pit stop, what would otherwise be an internal activity (putting the tire on the existing rim) has been converted to an external activity, at the expense of having an extra rim. Another example is the use of a kit of materials, tools, parts, etc., that eliminate searching and selecting. A large part of internal operations that is usually considered separately is measuring and adjustments, to assure that the value creating operation produces an output that meets specification. By standardization, these operations can be eliminated.

Simplify and automate all remaining internal operations to speed them up: Further setup reduction allows value creation to more closely match demand economically, but safety, repeatability, and ergonomics must not be sacrificed. This calls for simplification and automation of tasks such as fastening, moving, data entry, and communication. Special tools and programs, codes and signs, design for changeover, and other similar approaches are utilized to gain speed.

It is, of course, also necessary to apply some of these techniques and technologies to the external setup. With shorter setups, and shorter periods of value creation between setups, external operations must be carried out in less time. This means better organization, simplification, mistake proofing, and automation.

Shingo's insight: separate internal end external to reduce setup by half

Let’s look at what Shingo discovered about setting up stamping presses. He spent time observing the setup process, and had the great insight that there were two kinds of activities taking place: Those that required the press to be stopped in order to be carried out, and those that didn’t. Both were done while the was press stopped.

Typical activities that require the press to be stopped: unbolting the die, pulling it out of the press, bolting the die in place.

Typical activities that can be done while the press is running: finding the next die, and moving it next to the press, finding wrenches, and other tools, placing the new material next to the die.

Shingo called the first kind of activity “internal”, the second kind, “external”.

He found that typically, half the activites done while the press was stopped were external. As an industrial engineer, he realized that a bit of planning would yield a significant reduction in setup time. If the tasks that did not require the press to be stopped were done while the press was running, then the changeover would take half the time.

There are lots of reasons why internal and external elements are mixed – only one person is doing the setup, and there is no time to prepare, the next job isn’t chosen until the current run is finished, running the press and setting it up are separate jobs, and the setters wait until the run is finished, there is no pressure to set up quickly, the amount of time for setup is part of the schedule, so there is no need to rush, no one believes that setup can be reduced, everyone involved in the setup appear to be working as hard as they can, everyone hates setup, and want to stretch the run until the end of their shift.

The question that remains, once Shingo’s insight is understood, is how to put in place countermeasures to these reasons for long setups.