What is SMED?
SMED — Single-Minute Exchange of Die — is a lean manufacturing methodology developed by Shigeo Shingo that systematically reduces the time required to switch a production line or machine from running one product to running another. The term "single-minute" does not mean one minute — it means changeover time measured in single digits, i.e. fewer than 10 minutes.
SMED is the process of reducing changeover time — the elapsed time between the last good part of one production run and the first good part of the next — to under 10 minutes, through a structured separation and conversion of setup activities.
Changeover time is pure downtime: the machine is stopped, no product is being made, and value is haemorrhaging. In traditional manufacturing, changeovers could consume hours or even days, forcing factories to run large batch sizes to amortise setup costs — creating massive inventory, inflexibility, and waste. SMED attacks this directly. By reducing changeover time dramatically, factories can run smaller batches, respond faster to customer demand, reduce inventory, and improve OEE — all simultaneously.
The most dangerous kind of waste is the waste we do not recognise. Changeover time was for decades invisible — SMED made it visible and eliminable.
— Shigeo Shingo, Creator of SMED & co-developer of TPSHistory & Origins
SMED was not invented in a laboratory — it was forged on the factory floor through years of stubborn observation, experimentation, and relentless improvement by one of manufacturing's most creative minds.
Shigeo Shingo was first called to Toyo Kogyo's Hiroshima plant to address a bottleneck on a large body-moulding press. By studying and reorganising the setup process, he achieved an early reduction in changeover time — planting the seed for what would become SMED.
The defining moment of SMED's birth came at Toyota when Shingo was challenged to reduce the changeover time on a massive 1,000-tonne stamping press from four hours. Through rigorous analysis and applying his emerging internal/external framework, he reduced it to an astonishing three minutes — proving the method could deliver order-of-magnitude improvements even on the heaviest equipment.
Shingo published his landmark book A Revolution in Manufacturing: The SMED System in Japanese in 1983. The 1985 English translation brought SMED to Western manufacturers, where it was quickly adopted as a cornerstone of Lean manufacturing and the Toyota Production System worldwide.
SMED became embedded in Lean manufacturing programmes globally, taught in every Six Sigma curriculum, and cited in IATF 16949 and ISO standards. Today, digital tools — video analysis software, IoT-enabled setup timing, and AR-assisted changeover guides — are extending SMED into Industry 4.0 environments.
Core Concepts: Internal vs External Setup
The entire intellectual foundation of SMED rests on one critical distinction: the difference between Internal and External setup activities. Understanding and exploiting this distinction is where virtually all changeover time reduction comes from.
- Removing and mounting dies, moulds, or tooling
- Adjusting machine parameters and settings
- Centring, aligning, and locking fixtures
- First-off inspection and trial runs
- Bolting, clamping, and tightening fasteners
- Fetching and preparing next-run tooling and dies
- Pre-heating moulds or conditioning materials
- Staging materials, components, and consumables
- Preparing documentation and setup sheets
- Pre-setting tool dimensions on offline equipment
In most factories before SMED, 50–70% of internal setup time consists of activities that are actually external — they are performed while the machine is stopped simply because no one has ever separated and reorganised them. The first and highest-impact SMED intervention is almost always to identify this hidden external work and move it outside the machine stoppage window.
The 3 SMED Stages
Shingo structured SMED as a progression through three conceptual stages — each delivering a quantum of improvement. Teams move through these stages sequentially, though in practice all three are pursued simultaneously once the methodology is understood.
Study the current changeover in detail. Identify which activities are truly internal (require machine stoppage) and which are currently being done internally but could be external. Simply moving misclassified activities outside the stoppage window typically reduces changeover time by 30–50%.
Engineer solutions that allow genuinely internal activities to be performed externally. This may involve pre-setting tools offline, standardising heights and datums, or using intermediate jigs. This stage requires creative problem-solving and often delivers a further 25–40% reduction.
Eliminate waste within both internal and external activities — removing adjustments, replacing bolts with clamps, standardising component dimensions, parallelising tasks across a team, and mistake-proofing setups. This stage pursues perfection and may reduce remaining time by another 50%.
The 8-Step SMED Implementation Process
SMED is most effective when applied through a structured, disciplined process. These eight steps guide teams from initial observation all the way to standardised, sustained changeover excellence.
Film the entire changeover from start (last good part of old run) to finish (first good part of new run). Use a fixed camera capturing the full work area plus a handheld camera for close-up detail. Do not describe what should happen — record exactly what does happen. This step is non-negotiable: without an accurate baseline recording, all subsequent analysis is guesswork.
Review the video with a cross-functional team. List every discrete activity performed during the changeover and record its duration. Break activities down to the finest practical granularity — individual tool retrievals, walks, adjustments, and waits all deserve their own row. This produces a detailed activity log, typically 30–80 line items for a complex changeover.
For each activity on the log, classify it as Internal (machine must be stopped) or External (can be done while machine runs). Mark any currently-internal activities that are actually external — these are the low-hanging fruit. In most factories, 30–50% of logged internal activities turn out to be misclassified external work.
Reorganise the changeover sequence so all confirmed external activities happen before the machine stops (pre-set) or after production restarts (post-set). Create a new changeover procedure and checklist reflecting this. No hardware changes required — this step is pure procedure and organisation, yet typically delivers 30–50% time reduction immediately.
Engineer solutions that allow genuinely internal activities to be performed externally. Common examples: offline tool presetting (tools are measured and adjusted on a presetter, not on the machine); intermediate jigs that allow the new tool to be aligned offline before mounting; standardised datums that eliminate in-machine adjustment entirely. This stage requires investment in fixtures, presetters, or jig designs but delivers substantial further reductions.
Attack waste within both internal and external activities. Key techniques: Replace bolts with clamps or quick-release fasteners — every bolt turn is waste. Standardise tooling dimensions to eliminate height adjustments. Parallelise tasks by using two operators instead of one on critical-path activities. Eliminate trial runs and adjustments through precision presetting and standardised datums — the first part should always be a good part.
Once the improved changeover sequence has been validated and timed, capture it as the new standard. Create visual standard operating procedures (with photos or diagrams), produce a changeover checklist, and train all operators and team leaders on the new method. Standardisation ensures the gains are repeatable by any operator on any shift — not just by the engineer who designed the improvement.
Track changeover time as a formal production KPI — logged for every changeover event. Set targets and review trends in daily/weekly management meetings. Conduct regular changeover audits (re-film every 3–6 months) to verify compliance with the new standard and identify new improvement opportunities. SMED is not a project that ends — it is a discipline that compounds. Once single-digit changeover is achieved, the team often discovers further improvement through deeper analysis and technology investment.
Tools, Techniques & Principles
SMED is supported by a set of complementary tools and design principles. Mastering these accelerates setup reduction beyond what procedure changes alone can achieve.
Filming changeovers reveals wasted motion, waiting, and misclassified activities invisible to the naked eye during a live observation.
Replacing traditional bolts with cam clamps, toggle clamps, or bayonet fasteners reduces fastening time from minutes to seconds per joint.
Offline tool presetting stations measure and adjust tools to exact specifications before they are needed, eliminating in-machine adjustment entirely.
Standardising tool mounting heights, reference points, and clamping positions across the tooling range eliminates adjustment as a changeover activity.
Error-proofing changeover fixtures and tooling so that incorrect assembly is physically impossible — eliminating trial runs and first-off adjustments.
Structured pre-run and post-run checklists ensure all external setup activities are completed before the machine stops — every time, by every operator.
Using two operators with a precisely choreographed sequence to perform changeover activities simultaneously, halving the critical-path time.
Visual tool storage with dedicated locations for changeover tools eliminates searching and retrieval time — all tools staged and ready before the machine stops.
| Machine / Process | Industry | Before SMED | After SMED | Reduction |
|---|---|---|---|---|
| 1,000-tonne Stamping Press | Automotive (Toyota) | 240 min | 3 min | 98.75% |
| Injection Moulding Machine | Plastics Manufacturing | 180 min | 12 min | 93.3% |
| CNC Machining Centre | Aerospace / Precision | 90 min | 8 min | 91.1% |
| Tablet Press (Pharma) | Pharmaceutical | 4 hrs | 22 min | 90.8% |
| Filling & Labelling Line | FMCG / Beverage | 120 min | 14 min | 88.3% |
Use OEE data and Pareto analysis to identify the machine where long changeover times have the greatest impact on production efficiency, customer lead time, or inventory levels. Start with one machine — not the whole plant.
Record the complete changeover. Create the detailed activity log. Measure and baseline the current changeover time. Engage the operators as co-investigators, not subjects of observation.
Apply internal/external classification. Calculate potential time savings from reclassification alone. Identify which Stage II and III improvements are feasible given available budget and technology.
Run a SMED Kaizen event (typically 3–5 days) to implement and test all three stages. Measure after each stage. Film the improved changeover and compare against baseline.
Publish the new SOP. Train all operators. Add changeover time to the shift KPI board. Set an audit schedule. Use the capacity gained to justify expanding SMED to the next machine.
- Dramatically increases machine availability (OEE)
- Enables smaller batch sizes and mixed production
- Reduces finished goods and WIP inventory
- Shortens customer lead times and improves responsiveness
- Reduces risk of defects from complex long setups
- Frees capacity without capital equipment purchase
- Starting without filming the baseline — guessing the problems
- Applying SMED to the wrong machine first
- Stopping at Stage I and not pursuing Stages II and III
- Not involving operators in the analysis — losing buy-in
- Failing to standardise — gains erode within weeks
- Measuring only the "best" changeover, not average time
Industry Applications
SMED originated in heavy metal stamping but has proven universally applicable — wherever a machine or process switches between products or configurations, SMED delivers results.
The birthplace of SMED. Used on stamping presses, welding jigs, assembly fixtures, and paint booths across every major OEM and tier supplier.
Reduces GMP-compliant tablet press, capsule filling, and packaging line changeovers — where each setup requires documentation, cleaning, and validation.
Applied to filling lines, packaging equipment, and baking ovens. Cleaning-in-place (CIP) integration makes SMED complex but extremely high-value.
Plate changes and ink changeovers are classic SMED targets — where a 2-hour colour-change setup can be reduced to under 15 minutes with parallel teams and preloaded plates.
Offline presetting, pallet changers, and modular fixturing systems deliver single-digit changeovers on multi-axis machining centres producing high-mix, low-volume parts.
SMED principles applied to operating theatre turnovers and ICU bed changeovers — reducing downtime between cases and improving asset utilisation in high-pressure environments.
Summary
SMED is one of the most elegant and impactful tools in the Lean toolkit. Its genius lies in reframing changeover time not as an inevitable cost of doing business, but as a category of pure waste — measurable, separable, and systematically reducible. Every minute shaved from a changeover is a minute returned to production, a unit of inventory no longer needed in the warehouse, and a step towards the ultimate Lean goal: making exactly what the customer wants, exactly when they want it.
Key Takeaway
Mastering SMED means accepting that long changeovers are not a technical constraint — they are an organisational habit. The equipment that takes four hours to change over today is the same equipment that will change over in eight minutes after a disciplined SMED programme. The difference is not the machine — it is the method, the sequence, the fixtures, and the standards. Shingo proved it on a 1,000-tonne press in 1969. The same logic applies to every machine in every factory in the world, right now.
* 94% average time saving based on Toyota 1,000-tonne press case study (Shingo, 1985). Individual results vary by machine type, complexity, and starting baseline.
