Cycle-Time Reduction & SMED

Cycle-Time Reduction & SMED (CSSGB BoK V.C.2 — Analyze)

Quick Answer: Cut cycle time and lead time with continuous flow and smaller lots. Long changeovers force large batches—SMED (single-minute exchange of dies) systematically reduces setup time so mixed, frequent runs become economical.

BoK V.C.2 is at the Analyze level: diagnose why cycle time and changeovers are long, and structure SMED improvements—not merely name the acronym.

Cycle Time vs Related Metrics

If station cycle time exceeds takt, demand cannot be met without overtime, extra capacity, or rebalance. If cycle time is fine but lead time is long, waste is often waiting, batching, and inventory—not only slow operators.

$\text{Takt time} = \frac{\text{Available production time}}{\text{Customer demand}}$

Cycle-time reduction projects often rebalance work content, remove NVA steps, improve machine uptime, cut motion, and reduce setup so lot sizes can fall.

Continuous Flow

Continuous flow (ideally one-piece flow) means each unit moves to the next step with little or no queue. Benefits:

• Shorter lead time and less WIP.
• Faster defect feedback (problems surface immediately).
• Easier balancing to takt.
• Less space for storage.

Obstacles to flow: large batch policies, long setups, shared equipment, poor layout, unbalanced stations, and quality issues that force sorting piles.

When continuous flow is not yet possible, use pull with small supermarkets between processes as a bridge, while attacking the barriers (especially changeover time).

Why Setup Reduction Enables Flow

Economic lot size thinking says: if changeover is expensive (long), run large lots of one product to amortize the setup. Large lots → inventory waste, long lead times, delayed defects, and poor flexibility to demand mix.

If changeover drops from 2 hours to 10 minutes, the “penalty” for switching products collapses. You can run smaller lots, closer to true demand, and approach continuous flow or mixed-model production.

Simple illustration

A press runs three part numbers. Daily demand is mixed. Changeover is 90 minutes; management insists on weekly campaigns of each part → multi-day inventory. After SMED, changeover is 12 minutes; the cell runs each part daily in small lots. WIP and response time drop even if pure machine cutting speed is unchanged.

SMED — Single-Minute Exchange of Dies

SMED, developed by Shigeo Shingo, is a structured method to reduce changeover toward single-digit minutes (the name is aspirational—not every setup becomes under 10 minutes, but the methodology still applies).

Core distinction: internal vs external setup

Key insight: much “internal” time is actually external work done at the wrong time (searching for tools after the machine stops).

SMED stages (classic)

1. Observe and document the current changeover (video is ideal). List every element and time it.
2. Separate internal from external work. Move everything possible to external (prepare next die while running).
3. Convert remaining internal steps to external where technology allows (quick clamps, standardized interfaces, preset settings).
4. Streamline remaining internal work: parallel tasks with two people, functional clamps instead of bolts, one-turn fasteners, intermediate jigs, elimination of adjustments (centerlines, locators).
5. Standardize the new changeover procedure; train; audit; continue kaizen.

Typical SMED tactics

• Shadow boards and kits staged at point of use (5S + external setup).
• Quick-release clamps instead of many bolts.
• Standardized die heights and locating pins.
• Centerline settings and first-piece checklists reduced to essential checks.
• Color coding and poka-yoke so dies only mount one way.
• Parallel work (two technicians during internal window).

Metrics to track

• Total changeover time (last good A → first good B).
• Internal vs external minutes.
• First-piece scrap or adjustment loops after changeover.
• Resulting lot size, inventory days, and on-time mix flexibility.

Analyzing Cycle-Time Problems (Exam Lens)

When a stem describes long lead times and large batches:

1. Is cycle time at the bottleneck above takt? → capacity, rebalance, TOC, reduce process CT.
2. Is queue time dominant? → flow, pull, WIP limits.
3. Are changeovers long? → SMED; then reduce lot size.
4. Is work unbalanced? → line balance to takt; continuous flow cells.
5. Are defects and rework inflating effective cycle time? → quality/poka-yoke first.

Do not confuse SMED with TPM (total productive maintenance) or with general 5S alone—though 5S and standard work support SMED. SMED specifically targets setup/changeover.

Worked Conceptual Example

Video study of a packaging machine changeover (62 minutes):

• 18 min searching for change parts and tools after stop → move to external (kit staged).
• 12 min bolting with many fasteners → quick clamps (streamline internal).
• 10 min trial runs and adjustments → preset scales and locators (convert/streamline).
• Remaining pure internal mount time: ~15 min after improvements; total changeover target ~20 min including first good piece.

After SMED, the team cuts campaign length from one week to one day per SKU, inventory falls, and the value stream approaches continuous flow on high runners.

Green Belt Application Notes

• Use Analyze tools (time studies, spaghetti, bottleneck ID) to justify SMED and flow projects.
• Pilot the new changeover procedure on one machine before multi-line rollout (links to V.B).
• Update standard work for operators and document external checklists.
• In Control, track changeover time and first-pass yield after changeover as process metrics.

Exam tip: “last good to first good,” internal vs external, and converting internal → external are the highest-yield SMED facts. Continuous flow is the ideal state that short setups enable.