That gap matters. Changeovers that stretch to 60, 90, or 120 minutes limit how often a plant can switch products, push teams toward larger batches, and tie up cash in inventory nobody wants to hold.
This guide is written for US manufacturing and operations leaders, Lean teams, plant managers, and the frontline employees who actually perform changeovers. SMED gets mentioned constantly in Lean circles but is frequently misunderstood at the shop-floor level. Below, you'll find what it actually means, how the steps work in practice, where it applies, and where it isn't the right fix.
Key Takeaways
- SMED separates changeover tasks into internal work (machine stopped) and external work (done while running)
- Aim for single-digit-minute changeovers where practical, not a universal 10-minute rule
- Faster changeovers free capacity, make smaller batch sizes viable, and ease inventory pressure
- Pair SMED with 5S, standardized work, TPM, and Kaizen rather than treating it as a replacement
- Involve operators early and protect safety and quality throughout every changeover
What Is SMED and Why Is It Used?
SMED stands for Single-Minute Exchange of Die. It's a structured Lean method for cutting the time it takes to switch equipment, tooling, materials, or settings from one product or process to the next.
The word "changeover" covers more than the physical tool swap. It runs from the last good unit produced on the previous job at full speed to the first good unit produced on the next job, also at full speed. That includes shutdown, disassembly, cleaning, staging, installation, adjustment, and startup, plus any first-piece inspection along the way.
Where the Method Comes From
Industrial engineer Shigeo Shingo developed the SMED methodology while working with Japanese manufacturers, refining it over decades before it became closely associated with Toyota Production System-style quick changeover practices. The method spread because it gave teams a repeatable way to attack setup waste instead of just working faster or buying more equipment.
"Single-minute" is a target, not a guarantee. It means aiming for a single-digit-minute changeover, generally under 10 minutes, where the process allows it.
Some changeovers—especially those with extensive sanitation, allergen controls, or complex tooling—may never realistically hit that number. That's fine. The point is meaningful reduction, not an arbitrary number for its own sake.
How SMED Differs From Other Improvement Tools
SMED is easy to confuse with adjacent Lean and Six Sigma tools, but it has a specific focus:
- Not faster run speed — SMED doesn't make machines run faster during production
- Not preventive maintenance — though unreliable equipment can undermine SMED gains
- Not Six Sigma — Six Sigma targets variation and defects; SMED targets setup and changeover loss specifically
Why Organizations Use SMED
Shorter changeovers help operations teams:
- Free up equipment availability
- Make smaller batches economically viable
- Improve responsiveness to shifting demand
- Reduce pressure to overproduce just to justify a long setup
A 2022 case study in a ready-meal food factory documented the impact directly. After applying SMED principles, the plant cut total changeover time by nearly 30%, then brought it under 10 minutes with a line-hopping adjustment.
That work delivered an overall reduction of almost 68% and lifted OEE by 9 percentage points to 71%.
How SMED Works and Where It Applies
SMED follows a repeatable sequence: observe the current state, separate work by type, convert what you can, streamline what remains, and standardize the result—then run the cycle again.
Step 1: Study and Document the Current Changeover
You can't improve what you haven't measured. Teams typically use:
- Direct observation and video review (where site policy permits)
- Task-level timing for every activity, not just the "main" steps
- Operator interviews to surface hidden waiting or searching time
- Movement analysis to spot unnecessary walking or transport Baseline first. Capture total changeover time, internal (machine-stopped) time, variation between changeovers on the same job, and time to the first good unit. Without that baseline, you're guessing at improvement instead of proving it.
Step 2: Separate Internal and External Activities
Internal activities require the machine or process to be stopped. External activities can happen before shutdown, during the run, or after restart. Common external candidates include:
- Staging tools, fixtures, and materials near the point of use
- Pulling paperwork, recipes, or programs in advance
- Running pre-changeover quality checks on incoming materials Reserve downtime strictly for what genuinely can't happen any other way, usually disassembly, physical installation, and final adjustment.
Step 3: Convert Internal Work to External Work
This is where most of the time savings actually happen. Techniques include:
- Pre-assembly — building sub-assemblies or fixtures before the stop
- Duplicate or intermediate fixtures — swapping a pre-set fixture instead of adjusting one in place
- Preheating or presetting — getting temperature-sensitive tooling ready in advance
- Preloaded programs — queuing the next job's settings before the current run ends For example, a packaging line might preheat a sealing head on a secondary unit while the current run continues, then swap it in during the stop rather than heating cold. Every conversion still has to meet the same safety, quality, and traceability requirements as the original process. Speed gained by skipping a safety check isn't a real gain.
Step 4: Streamline and Standardize Remaining Internal Work
Whatever can't move offline still needs simplification:
- Quick-release fasteners instead of bolts requiring multiple turns
- Fixed guides and visual alignment marks to remove guesswork
- Standardized numerical settings instead of trial-and-error adjustment
- Parallel operator tasks so two people work simultaneously instead of sequentially Once the new sequence works, lock it in with visual work instructions, checklists, and clear role assignments. Without this step, gains erode within a few weeks as operators drift back to old habits.
Where SMED Is Applied
SMED originated in stamping and die-casting but now shows up across:
- Injection molding and machining
- Assembly and packaging lines
- Food and beverage processing
- Automotive production and maintenance interventions The same internal-versus-external logic extends beyond the factory floor. A 2022 peer-reviewed study treated operating-room turnover as a changeover: instrument prep and counting moved to a sterile trolley set up in parallel, and patient recovery shifted out of the OR. Researchers reported at least a 25% average reduction in changeover time for gynecology and general surgery cases, without new infrastructure or technology. Logistics, construction, and administrative work can use the same thinking whenever teams switch jobs or service conditions.
Key Factors for Successful SMED Implementation
Successful SMED depends on disciplined application: the right process, the right people, and safeguards for safety and quality.
Start With the Right Process
Not every changeover deserves this level of attention. Prioritize assets where changeovers are:
- Frequent
- Lengthy relative to run time
- Highly variable between occurrences
- Tied to a known constraint or bottleneck
Confirm this with operational data, not gut feel, before committing team time.
Involve the People Who Do the Work
Operators, maintenance, quality, engineering, and scheduling should all be part of observation, redesign, and testing. People who perform a changeover daily usually know exactly where the time disappears, long before a stopwatch confirms it. Skipping this step is one of the fastest ways to build a "solution" nobody actually follows.
Protect Safety and Quality First
Before changing sequence or removing a task, review:
- Lockout/tagout and machine guarding requirements
- Sanitation, allergen, or contamination controls
- First-piece approval and inspection requirements
- Traceability and restart procedures
A faster changeover that creates a safety incident or quality escape isn't an improvement.
Use Supporting Lean Practices Together
SMED rarely works in isolation. Pair it with supporting practices:
- 5S — keeps tools and materials organized for staging
- Standardized work — documents the best-known method and makes deviations visible
- TPM — keeps equipment reliable so the new sequence holds up shift after shift
- Kaizen or PDCA — refines the method over time instead of treating it as a one-off project
Measure Beyond the Stopwatch
Track more than elapsed changeover time:
| Metric | What it captures |
|---|---|
| Internal setup time | Time the machine is actually stopped |
| Changeover variation | Consistency between repeated changeovers |
| First-good-unit time | Speed to acceptable output after restart |
| OEE (Availability × Performance × Quality) | Broader equipment effectiveness impact |
| Defects and safety events | Whether speed gains came at a cost |
Leading North Advisors supports this work through Lean transformation consulting, practitioner-led training, and team coaching. The focus is connecting changeover gains to broader operational metrics and building the internal capability to sustain them.
Common Issues and When SMED May Not Be Appropriate
Misconceptions Worth Correcting
- SMED redesigns when setup tasks happen so changeovers shrink without rushing operators
- The single-minute target isn't universal or mandatory
- Reducing setup time should never bypass safety, sanitation, or regulatory controls
Where Implementations Go Wrong
- Measuring only the tool swap, missing preparation and startup losses entirely
- Redesigning without operator input, producing a method nobody actually uses
- Skipping standardization, letting the old process creep back within weeks
- Celebrating a one-time win without monitoring whether it holds
When to Look Elsewhere First
SMED isn't the default answer to every capacity problem. According to Lean Production, SMED becomes worth prioritizing when changeovers represent at least 20% of a process's lost productive time; below that, other losses deserve attention first. Consider a different starting point when:
- Changeovers are infrequent and low-impact for that specific asset
- Equipment itself is unstable or poorly maintained
- Product specifications are unclear or still changing
- The business doesn't actually need the added flexibility
- Quality, material supply, or scheduling issues dominate the real loss
Conclusion
SMED gives teams a structured way to attack changeover loss: observe it, separate it, convert what you can move offline, simplify what remains, and standardize the result so it sticks.
The value goes beyond faster setups. Done right, SMED supports smaller batches, better flow, more usable capacity, and steadier operations, but only when the underlying business problem actually calls for it.
Put SMED to work with a few non-negotiables:
- Start with data, not assumptions
- Involve the people running the changeover every day
- Protect safety and quality at every step
- Treat the standardized result as a starting point for the next improvement cycle, not a finish line
Frequently Asked Questions
What are the seven steps of SMED?
The seven steps are: (1) observe and measure the current changeover, (2) separate internal from external work, (3) convert internal work to external, (4) streamline remaining internal work, (5) reduce external effort, (6) standardize the new process, and (7) keep improving through repeated cycles.
What is the SMED concept?
SMED reduces changeover time by moving as much preparation as possible outside equipment downtime, then simplifying whatever work still requires the machine to be stopped. The goal is a single-digit-minute changeover where practical.
What is an example of SMED?
A packaging line pre-stages materials and pre-heats tooling while the current run continues, then uses standardized quick-release steps during the actual stop. The result is a shorter, more consistent changeover without added equipment.
Is SMED lean or six sigma?
SMED is primarily a Lean method focused on flow and setup-time reduction. Six Sigma tools can address variation or defects uncovered during a SMED project, but they don't define the method itself.
Is SMED part of TPM?
No, but the two complement each other closely. SMED reduces changeover loss, while TPM improves equipment reliability and availability. Used together, they usually lift OEE more than either method alone.