SMED

SMED (Single-Minute Exchange of Die) is a lean manufacturing methodology developed by Shigeo Shingo to reduce equipment changeover times to single-digit minutes (under ten minutes). The system operates by systematically categorizing setup tasks into internal activities (performed only when equipment is stopped) and external activities (performed while equipment runs). By separating, converting, and streamlining these tasks through standardized procedures and mechanical innovations, organizations drastically reduce downtime, facilitate smaller production batch sizes, and increase operational flexibility.

Preparing the press bay…

0/5 explored

Machine stopped

Goal: under 10 minutes

52 min

52m
Start the stopped clock
The baseline changeover begins before the next die, tools, or temperature are ready. All 52 minutes consume production time.
Stage the next die
The next die waits on a roller cart beside the press before the current run ends. Moving preparation outside the stop removes nine minutes.
Prepare tools and temperature
Tools are arranged and the next die is preheated while production continues. Together, external preparation cuts stopped time to 26 minutes.
Streamline the internal work
Removing and installing the die must remain internal. Roller transfer, quarter-turn clamps, and preset connections make that stopped work fast.
Prove the first good part
The changeover is complete only when the new setup makes a conforming part. This one does so after eight stopped minutes—a single-digit exchange.

Key facts

Developer
Shigeo Shingo
Development Period
1950–1970
Target Changeover Time
Under 10 minutes (single-digit minutes)
Core Setup Types
Internal setup and external setup
Implementation Stages
4 structured stages

By Matthew Savas — Founder of Kaizumi. Reviewed 30 August 2026.

SMED (Single-Minute Exchange of Die) is a lean methodology developed by Shigeo Shingo to reduce machine changeover times to under ten minutes, meaning a single-digit number of minutes. By separating external setup activities (performed while equipment runs) from internal setup activities (performed while stopped) and streamlining all remaining tasks, organizations eliminate downtime, enable smaller production batches, and improve flexibility. SMED provides the operational foundation for quick changeover programs and broader setup reduction efforts across manufacturing and service environments.

Try it yourself

Loading…

Origins and historical development

Shigeo Shingo developed the SMED methodology between 1950 and 1970 through his consulting work with Japanese industrial manufacturers, most notably Toyo Kogyo (Mazda) and Toyota Motor Corporation.

In 1950, Shingo conducted an efficiency study at Toyo Kogyo's Hiroshima plant, focusing on bottlenecks at 800-ton stamping presses. He observed that setup operations consisted of two distinct categories of work: operations performed while the press was stationary, and operations that could have been prepared while the press was running. By identifying and separating these activities, he cut press downtime substantially without buying new equipment.

In 1969, Toyota asked Shingo to attack the setup time of a 1,000-ton body-molding stamping press, which then took four hours. Six months of observation, classification, and mechanical modification brought it down to ninety minutes. Management then set a target of under ten minutes, and further work reduced the changeover to three minutes. Shingo formalized the principles derived from this achievement into the SMED system, establishing a repeatable approach for eliminating changeover waste.

Core concepts of internal and external setup

The SMED methodology rests on the distinction between two categories of setup tasks:

  • Internal setup activities: Tasks that can only be performed when the machine or process is completely stopped. Examples include removing an old die from a press bed, mounting a new cutting tool into a spindle, and connecting hydraulic lines to a replacement fixture.
  • External setup activities: Tasks that can be performed while the machine is operating and producing acceptable parts from the previous production run. Examples include retrieving raw materials from storage, pre-heating dies, gathering hand tools, reviewing production orders, and pre-setting blade dimensions on an offline staging bench.

In conventional, unoptimized operations, operators perform internal and external activities sequentially after stopping the machine. SMED shifts all possible preparatory and concluding tasks into the running cycle of the machine, reserving equipment stoppages strictly for actions that physically require an idle state.

The four-stage implementation methodology

Implementing SMED follows a structured four-stage process:

Stage 1: Document and analyze the baseline changeover

The team records the entire changeover process from the last good piece of the previous product run to the first verified good piece of the subsequent run. Video recording provides an objective record of every operator movement, machine state, tool retrieval, and adjustment step.

The team breaks the recording down into discrete task elements, noting the exact duration of each step, the physical distance traveled by operators, and the tools required. The team compiles these steps into a detailed waste analysis sheet, logging every delay caused by missing equipment, searching for fasteners, or waiting for quality inspection.

Stage 2: Separate internal and external setup activities

The team examines each documented task to determine whether it strictly requires the machine to stop, or if it can occur while the machine runs.

The primary objective in this stage is organizational discipline rather than mechanical modification. Tasks such as retrieving the next production order, locating the new tooling, transporting raw material to the workstation, and staging measuring gauges are moved outside the machine stoppage window.

In an illustrative press changeover requiring a 52-minute baseline stop, simply completing external tasks while the previous run is still underway can reduce stopped machine time to 26 minutes.

Stage 3: Convert internal setup to external setup

The team examines all remaining internal setup tasks to find methods for converting them into external activities. This stage often involves re-engineering tooling, adding pre-heating systems, or standardizing intermediate holding fixtures.

Techniques used in this stage include:

  • Pre-heating dies and molds using offline heating stations so they reach operating temperature prior to mounting.
  • Utilizing standardized mounting plates and sub-bolsters so tool dimensions are uniform across different machines, allowing pre-assembly offline.
  • Using duplicate tool holders and pre-setting cutting inserts using optical tool setters while the machine runs.
  • Standardizing centering and locating dimensions across varied tooling using locating pins and reference edges.

Stage 4: Streamline all remaining setup activities

The team systematically optimizes all remaining internal tasks and external processes to minimize total effort and duration.

For internal tasks, the focus centers on parallel operations, functional clamping, and the elimination of adjustments. For external tasks, the focus centers on standardized storage, dedicated tool carts, visual staging squares, and streamlined material handling routes.

Applying these streamlining measures can reduce the remaining internal downtime of a 26-minute changeover down to eight minutes, bringing the total stoppage well within the single-digit minute threshold.

Engineering techniques for setup reduction

SMED relies on several mechanical and organizational design principles to compress setup duration:

Elimination of threaded fasteners

Standard threaded bolts require multiple turns to secure or release, consuming valuable internal time. SMED replaces standard threads with functional clamping mechanisms that engage or disengage in a single motion:

  • U-shaped washers that slide under a partially loosened bolt head, requiring only a half-turn of the bolt to release.
  • Pear-shaped mounting holes where a bolt head passes through the large diameter and locks into the narrow slot with a partial turn.
  • Cam locks, wedge clamps, and toggle clamps that secure dies with a single lever stroke.
  • Pneumatic, hydraulic, and magnetic clamping systems that secure large tooling across multiple points simultaneously at the press of a button.

Standardization of tool dimensions and clamping heights

When tooling designs vary in base thickness, clamping height, or overall width, operators must adjust clamps, change packing blocks, and alter machine stroke limits during changeovers. Standardizing these external physical dimensions allows all tooling to use identical clamps at fixed heights, eliminating manual repositioning.

Elimination of adjustments and calibration

Adjustments account for a substantial portion of changeover duration and introduce process variability. Shigeo Shingo distinguished between settings and adjustments: a setting establishes a fixed numerical position, whereas an adjustment involves trial-and-error positioning accompanied by test runs.

SMED eliminates adjustments through mechanical positioning aids:

  • Fixed location pins, dowels, and positive stops that position tooling instantly without measurement.
  • Calibrated reference scales, digital readouts, and marked gauge blocks that replace manual tape measures and calipers.
  • Stepped stops and multi-position rotary dials with preset physical detents corresponding to specific product codes.

Parallel operations

When a machine changeover requires tasks on multiple sides of large equipment, a single operator must walk back and forth repeatedly. Dividing the internal tasks between two or more operators working in synchronized parallel paths eliminates redundant movement. Standard work combination sheets define the exact sequence of movements for each operator to avoid interference and maintain safety.

Impact on batch size and production flow

Setup duration directly dictates the minimum batch size an organization can produce economically. Traditional batch-and-queue manufacturing models determine production quantities using the Economic Order Quantity calculation, which balances changeover costs against inventory holding costs. When changeovers require multiple hours, large production runs are required to dilute the fixed downtime over many units.

Large batches generate long production lead times, high work-in-process inventory, delayed defect detection, and poor responsiveness to customer demand shifts.

Reducing setup time through SMED changes this calculation. When changeover time approaches zero, the economic penalty of changing from one product to another is eliminated. This allows production facilities to lower batch sizes toward single-piece flow and run mixed-model production sequences.

SMED enables the calculation of every part every interval (EPEI), also known as the production cycle frequency. To calculate the available time for changeovers within a shift, subtract total operating run time from total available shift time. To determine the EPEI, divide total changeover time required for all product variants by the available daily changeover time. As SMED compresses individual changeover times, the overall interval shrinks, allowing facilities to produce every product variant every day or every shift.

Practical applications and examples

SMED principles apply across various sectors outside traditional metal stamping:

Industrial stamping and injection molding

A metal fabrication plant operating heavy stamping presses evaluated a baseline tool change that took 4 hours. The team found that 60% of the duration was external work performed while the press was idle (locating dies, finding tools, and filling out paperwork). By staging tools beforehand and installing quick-release clamps, the changeover time was reduced from 4 hours to 12 minutes.

Healthcare operating room turnover

Surgical suites experience lost operational capacity during room turnovers between sequential patient procedures. Applying SMED involves categorizing room preparation steps into internal tasks (procedures requiring the patient to be in the room) and external tasks (preparation that can occur prior to patient arrival). By shifting non-operative tasks, such as room cleaning, instrument restocking, and patient preparation, into parallel external workflows, room turnover time was reduced from 45 minutes to 15 minutes.

Software deployment and IT systems

Database migrations and system updates often require scheduled maintenance windows where customer applications are taken offline. Applying SMED concepts involves separating tasks that require server downtime from tasks that can run in the background on parallel infrastructure. By pre-validating data templates, automating script checks, and staging resources in advance, scheduled system conversion windows were reduced from one hour to 10 minutes.

Sustaining setup improvements

Maintaining single-digit changeover times requires ongoing process control and standardized work:

  • Changeover standard operating procedures: Document the precise sequence, operator assignments, and tools required for each step.
  • Visual staging carts: Equip mobile carts with dedicated cutouts for every tool, fixture, and fastener required for the specific changeover, ensuring missing items are immediately visible.
  • Regular video audits: Periodically record and review changeovers against documented standard work to detect drift, identify newly emerging delays, and train new personnel.
  • Total productive maintenance: Keep tooling, clamping surfaces, locating pins, and quick-disconnect fittings cleaned and lubricated to prevent mechanical sticking or alignment failures during changeovers.

Frequently asked questions

How are internal setup tasks converted into external setup tasks in SMED?
Teams convert internal tasks by modifying tooling and procedures so preparatory work happens while the machine is still running the previous production run. Common methods include pre-heating dies in offline warming stations, pre-measuring cutting inserts in duplicate tool holders using optical setters, and standardizing mounting plates so tooling can be assembled in advance.
What is the difference between a setting and an adjustment in SMED?
In SMED methodology, a setting establishes a fixed numerical position, whereas an adjustment relies on trial-and-error positioning accompanied by test runs. SMED eliminates adjustments by replacing manual positioning with positive mechanical stops, calibrated reference scales, locating pins, and preset physical detents.
Why does SMED emphasize replacing standard threaded fasteners?
Threaded bolts require multiple turns to loosen and tighten, causing idle machine downtime during internal setup steps. SMED replaces standard bolts with functional clamping mechanisms that engage in a single motion, such as U-shaped washers, pear-shaped mounting holes, cam locks, and magnetic or pneumatic clamps.
How does SMED affect the Every Part Every Interval (EPEI) calculation?
EPEI determines production cycle frequency by dividing total changeover time required for all product variants by the available daily changeover time. Compressing individual changeover times with SMED shrinks the overall interval, enabling facilities to run smaller batches and produce every product variant every day or shift.
How are SMED principles applied outside traditional manufacturing?
SMED applies to any operational turnover by separating tasks requiring equipment stoppage from tasks that can run in the background. In hospital surgical suites, shifting room cleaning and instrument staging outside patient occupancy cuts room turnover from 45 minutes to 15 minutes. In IT environments, pre-validating data and staging updates on parallel infrastructure cuts scheduled server downtime windows from one hour to 10 minutes.

Matthew Savas — Founder of Kaizumi. Published 1 January 2025, reviewed 30 August 2026.