Applying SMED To Maintenance

If boosting available productive capacity and increasing wrench time has some appeal for your team, explore this fast-lane approach that the production side of the house has been using.

Shigeo Shingo (of Shingo Prize fame) is the inventor of the Single Minute Exchange of Dies (SMED) process. Some of his changeover accomplishments were so revolutionary at the time that they were thought to be impossible, even by visitors who watched an actual die change take place right in front of them. Today, however, many of these techniques are standard practice in manufacturing, and they should become standard practice in maintenance as well.

As in the application of various other lean principles, SMED allows operations to increase throughput and reduce WIP inventory. In the same way, applying SMED principles can allow maintenance to boost the amount of productive capacity available and increase the use of precious labor hours (wrench time).

Rev up
To get started in applying the SMED process, pick a PM or other procedure that fits one of the following criteria:

  • It takes a long time to perform.
  • It occurs on several pieces of equipment in the plant.
  • It occurs frequently.
  • It is on bottleneck production equipment.

Using these criteria will allow your solution to provide the biggest bang for the buck. Keep in mind that the goal of SMED is to reduce the amount of time that the equipment is down and still be able to carry out necessary maintenance tasks. The activities listed here are critical steps in the improvement process:

  1. Document the "as is process."
  2. Move internal (static) work to external (dynamic) work.
  3. Convert internal work to external work.
  4. Create a pre-PM plan.
  5. Simplify internal work—standardize.
  6. Verify PM quality.
  7. Complete post-PM activities.

The following principles are being applied:

  • To the maximum extent possible, maintenance should be done during scheduled work periods.
  • Maintenance should be performed on a schedule that is predictable and known by operations.

Change lanes
Let's examine what goes on in our operations when we need a maintenance task performed. We'll first review planned work, then later look briefly at how this technique applies to urgent or emergency work.

A planned job is in the backlog or a job to be scheduled. It is given—either electronically or on paper—to the crew supervisor, who assigns the work order to the craftsperson(s) who will then go through a typical work cycle, or, as shown in Fig. 1, a typical "as is process."


Fig. 1. Typical "as is process"


The result? Valued-added time of a craftsperson (wrench time) is about 25%! This is what is typically achieved using the normal planning and scheduling practices found in many plants across the country. The job assessment will be longer and there will be several trips to the storeroom. Special tools may have to be made or ordered or some special work-around applied. Documents may not be available—or may not be easily accessible. Overall, planned jobs should take far less time to execute than unplanned work, but this strategy is often not as effective as it needs to be. (Alas, this is a discussion for another time.) The unfortunate final result is that often things do not come up smoothly and operations wonders if the right things are actually being done. Consequently, there may be less of a push to schedule maintenance work during production periods in the future.

As noted before, there is a better way.

Document the "as is process"…
SMED is a technique where a person or team observes a job from beginning to end and fully documents the steps and the time used. It is very helpful to use a camcorder to record the work process so the improvement group can carefully study it later. Although a reliability engineer can do this work, such a person is probably better suited to improving MTBF (Mean Time Between Failure). A natural work group for this task includes the craftsperson(s), supervisor, planners and operators. The outcome from the improvement work will be executed by the craftsperson(s) and operators, with what has been learned then being applied to other work processes by the planner. This work will make the lives of the craftspersons better, so it must have their support. It also will increase the amount of work that a craftsperson can accomplish—therefore, it is critical to eliminate any resistance that non-involvement might create.

As you document the work-task procedure that is being followed, you will find that different people will perform the work differently. It is important to understand this so that the best procedure can be used as the starting point. It is also important to make sure all parties understand that once the new procedure is established, everyone is to perform it the same way. With everyone following the same procedure, errors will decline and reliability will increase.

Applying SMED principles can allow maintenance to boost the amount of productive capacity available and increase the use of precious labor hours (wrench time).

Move internal work to external work
Let's refer to internal work as "static work" and to external work as "dynamic work." Static work is just that—work done while the equipment is static or out of service. As team members review activities, they must classify each as either requiring the equipment to be out of service or that the work can be done while the equipment is still operating. Close attention must be given to the time required to perform these tasks. Later, when the entire PM procedure is rewritten, the activities must be grouped so that dynamic tasks—those done before the equipment is taken out of service—are completed prior to the equipment outage. Simple as it may seem, this step can greatly reduce elapsed production downtime.

Convert internal work to external work
Next, have your team look at the tasks that they have determined must be done while the equipment is out of service. With changes in procedures, tools, components used, etc., many of these tasks probably can be moved out of the static time period. This is a time for creativity and—particularly—a time to look at condition-based techniques that are available. Changes at this time typically require small investments, and the return can be enormous.

An example may be that a PM calls for a hydraulic oil change to be performed. An alternative is to take an oil sample, send it to the lab and let the test results tell you what maintenance is actually required. It may be necessary to install a sampling valve, as well as two quick disconnects to the reservoir to accomplish this change in procedure. By doing nothing or just doing simple filtering instead of scheduled oil changes, the savings in oil, waste-oil disposal, production downtime and maintenance labor can be substantial.

Create a pre-PM plan
Redesigning your PM procedure is key to gaining both efficiency and effectiveness. This is where the work structure is established. Work that is to be done prior to the static PM includes acquiring the parts, tools, technical information, necessary permits, etc. It will be necessary to perform the dynamic PM activities, like the previously referenced oil sampling, far enough in advance of the static work so that the results of the dynamic work can be used to modify the static PM task list. This shows that the PM program is a living, breathing process—and how a PM program can be truly effective. It also illustrates that effective planning is a key element of maintenance excellence, and that it will deliver capacity to the plant at the lowest overall cost.

Simplify internal work
With all of the static PM activities identified, including those that occasionally will be required as a result of dynamic PM findings, it is time to get the team to optimize the work. On large assets or production lines, doing so will require multiple craftspersons and operators working together to quickly perform the work.

The static work includes every activity from the shutting down of equipment by the operators, to lockout/tagout, to various maintenance tasks, to finally restarting the equipment and verifying that it is operating on the correct cycle time and meeting required quality. This process is similar to pit-stop action during a NASCAR race. So that total production downtime is minimized, it is important for the team doing the work design to carefully choreograph each participant's activities. It is also important to standardize this work. The design team should solicit input from other craftspersons and operators to write procedures that will be performed identically by different crews and shifts. Don't get too carried away with detail—at the same time, don't assume advanced skills or excellent memories either. (Fig. 2 depicts the SMED PM process.)


Fig. 2. The SMED PM process


Verify PM quality
As you can see, getting the operators engaged in the SMED process is very important. They can do many things—their two most important activities in this process are the proper shutdown of the equipment at the beginning, and properly bringing it back on line at the end of the PM work. The maintenance crew should remain at the work site until the operators, upon verifying the cycle time and quality, releases them. This builds the teamwork relationship that is essential in today's lean manufacturing environment.

Complete post-PM activities
The operators and maintenance team members will each have some post-PM activities to complete. For example, the operators may have to return small benches, equipment, tools or inventory to designated locations after production is back underway. The maintenance team will have to return new or used parts, tools, manuals and other items to their designated locations and fill out work-completion paperwork. All of this must be defined, usually in the form of standardized work, and is required by management. This work sets the stage for being successful when the next cycle of the respective PM is performed.

Validate effectiveness and zoom on
No doubt, some key performance indicators (KPIs) will have to be established to track how well the new process is working. It will be useful to track actual versus planned downtime and the percent of startups that occur without issues. Although these KPIs are operations-service metrics—not maintenance-efficiency metrics—they will help you keep everyone's collective eyes on the prize.

Try this new lean maintenance technique in a pilot area and once perfected, roll it—in the fast lane, of course—across your plant or corporation. So much seems to run faster these days; your maintenance activities can, too. MT

Wayne Vaughn, P.E., CMRP, former director of Maintenance for Harley-Davidson, is now a consultant for a wide range of industries. Telephone: (203) 550-1152; e-mail: This e-mail address is being protected from spambots. You need JavaScript enabled to view it