Printing Lean

Reducing setup time

November 2, 2007

Last issue we talked about Value Stream Mapping and how value stream maps (also known as "material and information flow diagrams") can help you "see" extraordinary amounts of waste inherent in your system. These are caused by:

Overproduction: Producing labels, forms, cartons, or signatures too soon or in a quantity greater than that desired by the customer;

Waiting: Includes operators who are standing idle next to a running machine waiting for a roll or lift to be completed. It also includes downtime waiting for materials, equipment breakdowns, or waiting for someone to sign off on the setup or run;

Transporting: Excessive movement of materials to and between people, equipment or storage locations;

Overprocessing: I call this one "overkill". Excessive or needless processing or reworking of a product, especially one that already meets customer requirements;

Movement: Any movement by a person that does not directly add value to the product. Having the right anilox in the press is value added. Walking around looking for the right anilox is waste;

Defects: What most of us think of when we hear the word waste: bad product or scrap;

Inventory: Having more raw materials, work-in-process, or finished goods than what is absolutely necessary to maintain continuous production or to provide what the customer wants.

We've also discussed how inventory is the "deadliest" waste of all, as excessive inventories are caused by or contribute to almost all of the other six wastes. Finished goods inventories are the "worst of the worst", because resources — labor and material — have been consumed to create a final product, but you, the manufacturer, have not been paid for these resources. Even worse, you now have to sit on these finished goods, in a sense becoming a warehouse for your customer(s), and you continue to consume resources to store and manage these inventories.

The reasons for excessive finished goods inventories are as long as your arm, but they almost always boil down to one: Once you've set up a job on press you want to run as much as you to can to avoid the hassle of setting up the same job just a few weeks or months down the road. You feel this way because you don't know, or haven't tried, to reduce your press changeover times that make this argument pointless.


SMED — or Single Minute Exchange of Dies — is a technique that has been used to successfully reduce changeover or makeready times on presses by as much as 90 percent. Think about that: Take every job you run, add up the cumulative setup times, and reduce that number by 90 percent. Now multiply that number by your average dollar per press hour and you start to realize the potential in continually practicing SMED. The value of this newly found press capacity is staggering.

Choose wisely

In a setup reduction exercise it is common for companies to target a press with the longest average setup time. That's fine, as long as that press is also a primary asset. If it's not — that is, if the press is a secondary asset and is not used very often — then you won't gain much by reducing its setup time. To put it another way, let's say for argument's sake that Press A has an average setup time of 50 minutes per job and Press B has an average setup time of 120 minutes. You've set a target of 25 minutes per job. Press B should be your target press then, right? Not necessarily. If Press B only runs one job per day you're only gaining 95 minutes of extra capacity, or about 395 hours per year. If Press A runs five jobs per day and you cut setup to 25 minutes per job you're gaining 125 minutes of extra capacity per day, or about 520 hours per year. Which would you rather have, 395 hours or 520?

As part of an overall Lean strategy you should look at where your capacity constraint lies — that should be the piece of equipment that you begin to work on. This constraint should have been identified during your Value Stream Mapping exercise (see last month's column).


As with any improvement project, you should follow the PDCA cycle when performing a SMED setup reduction event — Plan, Do, Check, Adjust (or Act).

Plan at least three weeks in advance in order to allow for production control to shift work off of the targeted piece of equipment; to allow for team members to make arrangements for child care, transportation or other personal issues (especially if you have team members from more than one shift); to allow for maintenance to prepare for supporting the team during the event (electrical, compressed air, water, machining, etc.); to allow shift supervisors to plan for coverage assignments to backfill team members who will now be "off the floor"; and to give yourself enough time to gather data that the team will need.

What, exactly, is setup?

SMED events follow a fairly standard protocol. The first thing that we must do is identify just what we mean by setup. Setup time is the time elapsed from the last good or saleable piece of product from the last job run until the first good or saleable piece of product runs on the next job. This is a crucial point: The clock starts running when the press stops running good product on the previous order, not when the operator "clocks in" to set up for the next job, and the clock doesn't stop until good product is running again.

Let's go to the videotape!

Now that we know what setup time is, we need to identify every task that takes place during a setup. This is impossible to do from memory, and standard operating procedures never provide enough detail to show every single step that occurs (especially steps that operators do "off the book"), so the best way to see what your current setup consists of is to capture a setup on video.

Position a camera — preferably one with a tape or disk large enough to capture an entire setup — as near as possible to the press without it being in the way of the operator(s). Do not follow the press operator around — this isn't an episode of COPS — and you'll just get in the way and possibly create a safety hazard. Instead, set the camera in an area that allows you to capture as much of the entire press length as possible. This will usually be from either the unwind or delivery end.

Turn the camera on and let it be. Have the operator(s) perform their normal setup routine. Do not turn off the camera for any reason until the setup is complete and good product is running. The objective is to capture everything that happens — good and not so good. You should video a setup during the three-week period leading up to the actual SMED event. Some people feel differently. They believe the data capture (video) should take place with the entire team present and should include having a team member timing each step with a stop watch while others spaghetti map every move or just simply stand there and observe. Personally I don't find much use for that and having too many people crowded around a press interrupts the operators' normal routine. That's not what you want to have happen. You also create a safety issue — too many cooks are in the kitchen.

Once taping is completed, the SMED team meets to review the video and document each step, including the time it takes to complete each step, so make sure the camera clock or recorder is running and is accurate. The team membership should include the operator(s) involved in the taping, or other operators who run identical equipment and can identify for the team what is happening on the tape, as well as stakeholders from processes or operations before and after the press operation: customer service, sales, material handling, finishing, shipping, maintenance, etc.

The team should have at least five but no more than eight members, and no more than half of the team should come from the target area. This will prevent the press department or group from "hijacking" the team by controlling the discussion and stifling creativity. Team members from pre- and post-press departments will be able to provide fresh ideas, and will also be able to see the impact that their departments have on press operations.

Internal vs. external

Once each step in a setup has been identified, the next part of the SMED process is to identify each step as either internal or external. Simply put, internal steps are those that cannot be performed while the press is running. Obvious examples include changing plates and tooling. Anything that is not an internal step is an external step and can be performed while the press is running, and it's these external steps that end up being the causes behind lengthy setups.

Examples of external steps include:

-    Getting materials for the next job (paper, inks, plates, etc.);
-    Reviewing paperwork;
-    Entering data on press sheets or into a computer;
-    Cleaning the press area;
-    Wrapping, strapping, packaging previously printed product.

There are many others. The question that you must ask for each step is: "Can this be done while the press is running?" For each internal step the team should brainstorm ways to convert as many as they can to external.

Once you've identified every internal and external step, add up the total internal and external times. Your cumulative internal time is your goal — your "ideal" time. You will usually find that this is at least 50 percent of your original setup time. If it's not, go back and review your internal steps. Are they really internal or can they be made external, or can they be minimized in any way?

A new better way

At this point the team will readily see that much of what was previously thought of as "as good as it gets" can, in fact, be better. You'll see operators disappearing off camera to find missing tools, materials or job information that was not provided, looking for someone to clarify a job instruction or to sign off on a first piece, or retrieving the same tool over and over again. All of these things are waste and should be eliminated. The team's task is to identify how to do this — how to get as close as possible to that ideal, internal time.

When a new, better setup procedure has been established, the team will need to perform a trial run using the new procedure. This trial should be videotaped the same way as before and from the same camera angle. This will allow the team to critique its procedure and make adjustments, if necessary.

Follow-through is critical

I have been involved in or have had oversight over more than 40 makeready events using SMED. Not once has a team failed to achieve at least a 50 percent reduction from its original baseline. Some have exceeded 90 percent. Needless to say, the team and the management staffs were shocked at these results.

After the team finalizes the new setup procedure it is up to the management team to sustain the gains. New work procedures may need to be written, new training materials may need to be developed, and some people may still need to be convinced that this new procedure is a better way. These tasks fall squarely on the leadership team and not on the SMED team. The SMED team's job is to show that it can be done — management's job is to make sure that it is done.

A common, fatal mistake that companies make when they perform a SMED event is that they believe that once the original event is over they've completed their mission — one and done as they say. Nothing could be further from the truth. Continual improvement requires, demands even, that you go back and keep working at reducing setup times. Today's best practice may not be tomorrow's. New technologies emerge; employees have new experiences and gain new skill sets; business conditions change. A Lean enterprise continually seeks to make improvements and is never satisfied with "good enough". Revisit your SMED target areas again and again and continually drive out waste from the process.

Tom Southworth is a business development manager with CONNSTEP, Connecticut's Manufacturing Extension Partnership. CONNSTEP's mission is to help Connecticut manufacturers apply advanced manufacturing techniques and management methodologies to become more competitive. He is a senior member of ASQ and is a Lean Bronze Certified Sensei by SME. His e-mail address is