SMED change-over analysis

How the SMED analysis is calculated

SMED (single-minute exchange of die) shortens a change-over by separating the work that needs the machine stopped from the work that can be done while it runs, moving as much as possible to the second kind, and then shortening what stays. The method is a common practice, not a standard; the categories and actions here are the usual ones.

Formulas

• Change-over time = the sum of the internal steps (machine stopped). External steps are work, but not downtime.
• After: an eliminated step is gone; a step converted to external moves to the external time with its duration (or the new one); a shortened step takes its new duration; a kept step stays, with a new duration if one is given.
• Saved per change-over = internal before − internal after; reduction = saved / internal before.
• Run time = demand × cycle time; free time = available time − run time: the time left for change-overs in one period.
• EPEI (every part every interval, in periods) = variants × change-over time / free time.
• Batch per variant = demand / variants × EPEI: equal demand for every variant is assumed. Change-overs that fit per period = free time / change-over time.

Worked example (the tool’s example)

1. Eleven of the twelve steps of the die change are internal: 12 + 8 + 6 + 10 + 5 + 12 + 7 + 9 + 15 + 10 + 4 = 98 min of stop; returning the old die (8 min) is already external.
2. Fetching the die, finding the tools and the paperwork (24 min) move to external; quick clamps, pins, couplings, shims and fewer trial parts shorten five steps; the fine adjustment is eliminated.
3. Internal time after: 2 + 10 + 5 + 8 + 3 + 2 + 6 = 36 min, saving 62 min per change-over, a reduction of 63.3 %.
4. 2400 pieces a day at 12 s take 480 min of the 960 available, leaving 480 min. With four variants EPEI = 4 × 98 / 480 = 0.82 days before and 4 × 36 / 480 = 0.30 days after; the batch of each variant drops from 490 to 180 pieces.

Common mistakes

• Timing the change-over from memory instead of observing it: steps like searching for tools are forgotten.
• Shortening internal steps before moving the external work out of the stop: the cheapest gain comes first from preparation.
• Counting external work as saved time: it still needs a person, only not a stopped machine.
• Ending the change-over at the first part instead of the first good part at full speed: trial runs and adjustments are part of it.
• Reading EPEI with unequal demands as exact: the batch shown is for an average variant.

Input and files

• Numbers may use a decimal comma or point. The type, category, action and new duration columns are found by their header names and may be left empty: a step is then internal, not categorised and kept.
• Excel export: the steps and a summary sheet whose totals, reduction, EPEI and batch sizes are formulas, so changing a duration, a type or the demand in Excel recalculates them.
• Print / PDF prints the results on one A4 landscape sheet; use “Save as PDF” in the print dialog. The blank observation sheet (fillable PDF or Excel) is for timing a change-over on the floor.
• Files: CSV, TSV and TXT are read as if pasted (UTF-8, UTF-16, otherwise Windows-1257). XLSX: choose the sheet and the columns or a cell range; formula cells use the value Excel last saved, dates are kept as text. Files up to 20 MB are read in this browser only and never uploaded.
• XLSX files are read with SheetJS Community Edition 0.20.3, Apache License 2.0 (license text at the end of this file).

Questions

What is the difference between internal and external work?

Internal work can only be done while the machine is stopped, such as taking out the old die. External work can be done while the machine still runs the previous order or already runs the next one, such as fetching the next die, preheating it or getting the tools ready. Only internal work adds to the change-over time.

Where does SMED start?

With watching and timing one real change-over step by step, ideally on video. Then mark each step internal or external, move everything possible to external, and only then shorten the internal steps (quick clamps, locating pins, standard heights, parallel work) and remove adjustments and trial runs. The blank observation sheet is made for the first part.

What is EPEI and why does it get shorter?

Every part every interval: how many periods (days, shifts) it takes to run every variant once. It is the number of change-overs in one round times the change-over time, divided by the time left after running the demand. A shorter change-over leaves the same free time for more change-overs, so every variant can be made more often in smaller batches, with less stock.

Does a shorter change-over always mean smaller batches?

Only if the saved time is used for more change-overs. It can also be used for more output, when the machine is a bottleneck. The tool shows the smaller batch it allows; whether to take it is a decision about stock against capacity.