SMED Guide: Faster Changeovers with Structured Checklists

Observation checklist

Use during actual changeovers to record sequence timing, identify waste, and capture improvement ideas.

Preparation & readiness checklist

Use before the changeover starts to confirm tools, materials, and roles are in place, reducing avoidable delays.

Sustainment & audit checklist

Use routinely after implementation to verify practice consistency, capture deviations, and drive continuous improvement.

1. What is SMED?

SMED stands for Single-Minute Exchange of Die. It is a Lean methodology for systematically reducing the time required to change a machine, production line, tool, product, or process from one setup to another.

The goal is not necessarily to achieve literally less than one minute, but to reduce changeover time to single-digit minutes where technically and economically feasible.

2. Why should we reduce changeover time?

Shorter changeovers can help organizations:

• Increase available production capacity
• Reduce planned downtime
• Improve equipment utilization
• Reduce production batch sizes
• Increase flexibility
• Reduce inventory and work-in-progress
• Respond faster to customer demand
• Reduce startup waste and scrap
• Improve production flow

3. Where should we start with SMED?

Start with one real changeover on one machine or process.

Choose a process where:

• Changeovers occur frequently
• Changeovers take significant time
• Downtime has a measurable business impact
• Operators experience recurring problems
• The team is motivated to improve

Avoid starting with the most complex process in the organization unless there is strong experience and support available.

4. How do we choose the right changeover for a SMED project?

Consider three factors:

Frequency × Duration × Business Impact

A changeover that takes two hours but happens once per year may be less attractive than a 30-minute changeover that happens several times every day. A good starting point is often a process with high frequency and significant cumulative downtime.

5. How do we measure changeover time?

First, clearly define the start and end points.

Start: Last good product from the previous production run

End: First stable good product from the new production run

This captures not only the physical equipment change but also adjustments, testing, startup losses, and stabilization.

6. Should we observe the changeover before making improvements?

Yes. The first step should be to observe the actual changeover as it happens. Do not rely only on standard procedures or interviews.

Record:

• Every activity
• Actual duration
• Waiting time
• Walking and movement
• Tool changes
• Adjustments
• Cleaning
• Quality checks
• Problems and interruptions

The objective is to understand the current state before trying to improve it.

7. What is the difference between internal and external setup?

This is the foundation of SMED.

Internal setup: Activities that require the machine or equipment to be stopped.

External setup: Activities that can be performed while the machine is still running.

The first major SMED opportunity is to identify activities currently performed during downtime that could instead be completed before the machine stops or after production resumes.

8. Does SMED mean every changeover must take less than 10 minutes?

No. "Single-minute" means the objective is to reach a single-digit number of minutes where feasible, but the appropriate target depends on the process.

Some processes may realistically achieve 5 minutes. Others may require 30 minutes due to technical, safety, quality, or regulatory constraints. The important principle is to systematically reduce unnecessary downtime, not to pursue an arbitrary number.

9. How much changeover time should we aim to reduce?

There is no universal target. A practical approach is to establish the current baseline and then set progressive targets.

Current: 90 minutes

→ First target: 60 minutes

→ Second target: 40 minutes

→ Future target: 20 minutes

The target should be based on observed losses and improvement potential rather than an arbitrary percentage.

10. Should we start by buying new equipment?

Usually, no. SMED should first focus on improving the existing process.

Start with:

1. Observe
2. Measure
3. Separate internal and external activities
4. Eliminate unnecessary work
5. Move activities outside downtime
6. Simplify the remaining work
7. Standardize the improved process

Only then consider investments such as quick-release tooling, automated systems, fixtures, or specialized equipment.

11. How important are operators in SMED?

Extremely important. Operators often know the practical problems that are not visible in production data. They know where tools are missing, which adjustments are difficult, and which steps consistently cause delays.

SMED should be done with operators, not to operators.

12. Should maintenance and engineering be involved?

Yes. Operators should provide practical process knowledge, while maintenance and engineering can help identify technical improvements such as:

• Quick-release mechanisms
• Standardized tooling
• Preset adjustments
• Improved fixtures
• Parallel operations
• Automation opportunities
• Error-proofing

The strongest SMED projects typically involve production + maintenance + engineering + quality.

13. What if the changeover process is different for every operator?

This is a common starting problem. If different operators use different methods, first identify the best-known method and document it.

Then standardize:

• Task sequence
• Tools
• Settings
• Responsibilities
• Expected times
• Quality checks

Once a standard exists, it becomes possible to measure deviations and improve the process systematically.

14. What are the most common sources of changeover waste?

Typical sources include:

• Searching for tools
• Waiting for maintenance
• Waiting for materials
• Unnecessary walking
• Cleaning delays
• Long tool removal and installation
• Manual adjustments
• Trial-and-error settings
• Repeated quality checks
• Missing documentation
• Poor communication
• Equipment that is difficult to adjust
• Lack of preparation before shutdown

The key is to observe and measure these losses instead of assuming where the problem is.

15. How do we know which activities should be improved first?

Prioritize activities based on:

Time impact × Frequency × Ease of improvement

High-impact activities that occur frequently and can be improved relatively easily are usually good first targets. You should also consider safety and quality. A small improvement that eliminates a recurring safety risk may be more valuable than a larger time saving.

16. Should changeover tasks be performed in parallel?

Yes, where safe and practical.

For example, instead of:

Task A → Task B → Task C → Task D

you may be able to organize:

Task A + Task B + Task C simultaneously → Task D

Parallel work can significantly reduce elapsed changeover time, but it requires good coordination, clear responsibilities, and appropriate safety controls.

17. How do we deal with adjustments and trial-and-error?

Treat adjustments as a separate source of changeover loss.

Look for ways to:

• Use preset values
• Add visual reference points
• Standardize settings
• Use scales or indicators
• Introduce digital recipes
• Improve tooling precision
• Use poka-yoke
• Prepare settings offline

The goal is to make the machine "right first time" rather than relying on repeated adjustments after startup.

18. How should we involve quality in SMED?

Quality should be involved from the beginning.

The objective is not simply to make the machine restart faster; it is to achieve the first good product faster.

Quality teams can help reduce:

• Startup scrap
• Repeated testing
• Waiting for approval
• Trial production
• Adjustment loops

The best SMED metric is therefore often not just changeover duration, but time to first stable good production.

19. What KPIs should we track?

Recommended indicators include:

• Average changeover time
• Median changeover time
• Best demonstrated changeover time
• Changeover time variation
• Percentage of changeovers meeting target
• Internal setup time
• External setup time
• Time spent on adjustments
• Startup scrap
• Time to first good product
• Number of changeover-related disruptions

For management, also consider measuring total changeover downtime per week or month, as this shows the overall business impact.

20. How do we prevent improvements from disappearing?

SMED improvements must become part of the standard operating system.

Use:

• Standard work
• Visual instructions
• Changeover checklists
• Operator training
• Regular observation
• Performance dashboards
• Periodic audits
• Refresher training
• Continuous improvement reviews

A successful SMED project is not complete when the target is achieved. It is complete when the new method becomes the normal method.

21. How often should we audit changeovers?

The frequency should depend on the importance and stability of the process.

For critical or high-frequency changeovers, consider weekly or monthly observations initially. Once performance is stable, reduce the frequency while maintaining periodic verification.

The audit should increase in frequency if:

• Changeover performance deteriorates
• New operators join the team
• Equipment changes
• Product mix changes
• New tooling is introduced
• Standards are repeatedly not followed

22. What should we do if the team says, "This is the way we've always done it"?

This is one of the most common barriers.

Avoid immediately challenging the team's experience. Instead, observe the actual process and ask:

"What would make this changeover easier, safer, or faster for you?"

Use real data and direct observation to identify opportunities. SMED works best when operators become owners of the improvement, rather than feeling that a new process is being imposed on them.

23. How long does a typical SMED improvement project take?

It depends on the complexity of the process.

A focused pilot can often be completed in a few days to several weeks, while major technical improvements may take longer.

A good approach is to start with a focused improvement event, demonstrate measurable results, and then expand the methodology to other machines and processes.

24. What should we do after the first SMED project?

Do not immediately launch dozens of projects.

Instead:

1. Document the results.
2. Standardize the improved process.
3. Train the team.
4. Verify that the improvement is sustained.
5. Capture lessons learned.
6. Select the next high-impact changeover.
7. Repeat the improvement cycle.

Over time, SMED can become part of the organization's broader Lean transformation system, connected with 5S, Standard Work, Visual Management, TPM, OEE, Kaizen, and Value Stream Mapping.