SMED is a tool used in the IMPROVE phase of a DMAIC project to improve existing set-up times enhancing the flow of products and alleviating a constraint. It can be used in a manufacturing or transactional environment.
At this point the team has determined that set-up time is a KPIV and improvements here will close the gap as described in the problem statement on the project contract.
Defining a "Set-up"
The time elapsed since the last good part (or last good transaction) was produced until the first good piece (next good transaction) is completed in the next production run.
Think about that entire statement for a few moments with the team.
There is much more time involved than just wrenching, inspection, and adjusting in a set-up. A "change-over" is not a set-up, it is a part of a set-up.
This includes the amount of time it takes to complete initial paperwork, removing the previous job, cleaning machine area, installing next job tools, running warm-up or trial pieces, getting them inspected, and finally running the first piece of the production run.
There may be brief moments of "run time" after some adjustments are made and then the machine is shut down again, then more adjustments made and the cycle continues. All of that time is classified as part of the total set-up time until the first good piece of the production run is made.
Speed offers competitive advantage over others especially if you line of work is produces a commodity product or common service.
While quality levels must be maintained, the organization that is most consistent (lowest variation in set-ups) and the fastest (lowest average set-up time) has an opportunity to improve cash flow, quote more competitively and offer customer higher delivery performance levels.
Whether it’s replacing a light bulb, making toast, assembling an engine, they are all repetitive tasks that can be streamlined to make life at home and work more value-added and efficient.
It is not necessarily working harder on every set-up for the duration of our work careers, it is about working very hard during the team's duration to permanently reduce set-up times and consistency...EVERY TIME. Create a smarter process to enable quicker set-ups. There are a few basic steps that a team can follow:
1) Identify set-ups through Pareto charts or other tools that display the vital few key improvement opportunities (inputs to the problem). Those that have both the highest amount of average time (mean) and the most variation are KPIV's.
2) Ensure the team is now expanded for this SMED portion to include operators, supervisors, line leaders, and other with a direct interest.
3) Video-tape the entire event if permissible. It can be broken down, analyzed and provide limitless mining opportunity and referenced if there is debate. Everyone should remain as natural as possible, act as if the camera is not on.
Recording starts when the last part of the preceding lot is complete and stops when the first good part of the next batch is produced. The camera rolls even if operator(s) leave the viewing area or can not be captured in the area (such as bathroom break, or to adjust in a area under a machine that is hard to view).
Ensure there is a way to maintain a charged battery or other method to capture the entire set-up. The camera must not be stopped when the operator leaves the machine for any set-up related operation.
4) Others should be on site taking notes of the process, identifying details that the video may not capture. Sometimes multiple tasks are occurring at once in different locations. Look for the low-hanging fruit and quick hits. This builds momentum.
The note-taker should capture tool specifics, sizes of adjustment pins, rails, bolts, wrenches, form numbers, smells, and other details the camera will not be able to detect.
6) Assign one person to doing a deep dive into one type of the 7-Wastes. It will be their job to expose and discuss all they can identify of one type of waste.
7) Watch the video tape and possibly break-out into groups and get more involved including those responsible for production, maintenance, quality control, and material handling.
8) KEY STEP: Determine which steps throughout the entire set-up are internal and external, and the amount of time of each.
INTERNAL - a step that can be done while the machine IS NOT running. Common examples (but not always):
EXTERNAL - a step that can be done while the machine IS running
9) First try, ELIMINATE steps.
10) Secondly, try to convert the INTERNAL steps to EXTERNAL steps for the obvious reason that the machine can run during those steps and operators must be disciplined to run in that manner (should be clearly documented in the standard work).
For example, a set-up may have involved a shut-down while the QC department inspected a critical part. Depending on their department tasks at that moment, this wait time could take 1-20 minutes. If it is found that the operator(s) very rarely have failures then the risk may be financially justified to run production during this time. The risk is scrapping parts if there is a failure.
Try cross-training the operator so they can make a better quality control decision themselves or enhance the QC feedback process.
11) REDUCE TIME in each element, internal or external, but put a priority on the internal elements. Advanced kitting, die packs, mold preparation, spare perishable parts, tugger routes, and other preparation are ways to reduce set-up time.
Think about a pit stop at a racetrack, 15 seconds is too much time. Yes, they have a lot of help but look at the standardization and readiness of all the items. A set-up is an "emergency", money is being lost when the machine is not making parts or employees are not performing a customer paid for service.
Think about SYNERGY.
Can the sum of two or more people get more accomplished than one person with their same combined time?
Sometimes the efforts or two or more working simultaneously can cut down a set-up by so much that it financially justifies that additional labor.
12) Standardize bolts, wrenches, forms, platforms across machines, lighting, oils, tools, workbenches, clamps, pins, dies, quick adjust handles, gages, and anything that will remove variation across the Families of Variation (shift-to-shift, operator-to-operator, machine-to-machine, part-to-part, etc.)
Settings should be used instead of "feel" or regular adjustments, using dial indicators help ensure safety, machine integrity, tool behavior consistency, and reduces more variation.
14) Conduct another intense 5S audit of the area.
15) Point-of-Use (POU) tooling. Everything should be organized and labeled with most frequently needed items on the operator or close by to reduce waste of motion and waiting, every second counts.
16) Spread the best practices across the system, operators, machines, and shifts.
Individual operators conducted setups as they were taught, often by different operators with varying methods over time and then they tacked on a few of their own ideas. This may result in consistency and being comfortable within oneself; however, it creates variation among (when compared to) other operators.
In other words, the operator is repeatable but the setups are not necessarily reproducible (see Gage R&R for explanation).
Sometime it is the "habits and procedures" that can be changed first, before any significant investment is needed.
Shift changes were not comfortable with set-ups or the preparation that was done on other shifts so the new shift operator will re-do it.
There is a general belief that longer set-ups result in better quality. While it is worth while taking the time to do a thorough and complete set-up, it is not time itself that results in better quality. Find those time consuming adjustments and make them quicker and standardize.
A common reason any shop will resist batch size reduction is because set-up are time consuming and operators will always be doing them. That is correct, batch size reduction CAN NOT start until downtime is eliminated, whether it is from SMED, product scheduling sequencing, or other methods.
To exaggerate in an attempt to make a point, if a set-up took only 15 seconds every time (such as race track pit stop) and was consistently that amount of time, then the operators would have much less resistance to doing them more often and running smaller batches.
More frequent changeovers or set-ups seems like a ridiculous concept to many traditional shops. While getting to truly one-piece flow may not be practical it is important to build economic production quantities.
Part of that formula involves downtime and a large component of downtime is set-up time.
More frequent changes allow for shorter production cycles of smaller batches (assuming the customer consumes them before the next batch behind it arrives).
Smaller batches that are consumed as they are finished will reduce amount of cash tied up in value stream. It also adds capacity if needed due to the extra time now available to take on new work. This absorbs fixed costs and creates more profit per piece as utilization of assets increases.
Quicker set-ups also allows for quicker response to hot customer requests or enhanced commitments to lead time that a customer may be just as or more concerned with than price.
If the smaller batches are just piling up together in the various WIP staging areas then the system is not flowing and it is actually costing more than running the larger batches.
In this case, you might as well run the larger single batch to reduce paperwork, inspections handling, because if they are piling up they are being run back to back most likely in the same manner as a large batch.
This module provides additional insight on Single-Minute Exchange of Dies. Click here to purchase the SMED module and view others that are available.
There is a direct correlation to batch size and set-up time. If the set-up time is reduced by 75%, then it is expected that batch sizes should be reduced in a similar amount.
Downstream or upstream operations MUST be considered. Producing in small batches may work fine at some operations but other operation, such as in a large furnace or oven, may be very difficult to financially justify due to all the breaks in production the smaller batches will generate.
Some set-ups take several days, some take a few minutes. The goal is to strive for a 90% improvement.
Work to shave off more time each trial. It won't all be possible at once and don't prepare to achieve this or expect the team to get this accomplished. Remember, it is called continuous improvement for a reason.
Take the necessary steps (safe and practical) to prove it can be done and the team's job is to control and sustain that level of performance.
Six Sigma Modules
The following presentations are available to download
Green Belt Program 1,000+ Slides
Cause & Effect Matrix
Central Limit Theorem
1-Way Anova Test
Correlation and Regression