In a garment factory, OEE fits machine-paced work such as knitting, automatic cutting and sewing automats. On a manual sewing line the operator sets the pace, so line efficiency based on SAM stays the main number. You can still calculate line OEE, using the line's proven rate as the ideal: line efficiency = OEE x balance ceiling.
Two different kinds of work under one roof
A knitting machine or an automatic cutter runs at a speed set by the machine. A single-needle operator stitches for short bursts and spends the rest of each cycle picking up, aligning and passing work.
OEE was built for the first kind. Garment plants already run the second kind on labor metrics, and they have good reasons for that.
| Area | Who sets the pace | OEE fit |
|---|---|---|
| Knitting, circular machines | Machine | Good. See textiles. |
| Automatic cutter, spreader | Machine and marker | Good, in meters or plies |
| Pocket setters, automats, embroidery heads | Machine, operator loads | Good |
| Whole sewing line | Slowest station | Useful next to line efficiency |
| One manual sewing operator | Person | Poor. Use individual efficiency against SAM. |
What differs on a sewing line
A line stop or a broken thread?
Line-level stops are the ones that stop output: style change, no cut panels, a failed machine at a station with no cover. A broken thread on one of 30 machines usually just thins the buffer, so it shows up as slow running.
Keep the threshold simple and write it down. Minor stops vs breakdowns helps you draw the line, and waiting for material covers the panel shortage.
SAM is not the ideal
SAM, or standard allowed minutes, is the usual time base in sewing. Abtew and co-authors (AUTEX Research Journal, 2020) describe it as a measure of a garment's work content, with the value made by adding allowances to basic time.
For OEE, take the rate the line has proven on this style with this headcount. It sets the ideal at the station that really limits the line, not at a perfectly balanced line that does not exist.
Count at the end of the line
Count garments that pass end-of-line inspection the first time. A garment repaired and passed on the second check is rework, and that counts against quality here even if it ships.
How line OEE and line efficiency connect
This is the part most factories want to know. The two numbers use different ideals, and one simple factor links them.
This holds when operators stay at the planned number, the minutes match scheduled time and both use the same good count. Many factories count every passed piece in line efficiency, repairs included, which lifts it a little, and garment line efficiency covers that formula in full.
Worked example: one sewing line, one day
Example: a polo line has 30 operators, a garment SAM of 20 minutes and 450 scheduled minutes. If it were perfectly balanced it could make 30 / 20 = 1.5 pieces a minute, but the best rate it has proven is 1.25 pieces a minute, or 0.8 minutes a piece.
Stops were 40 minutes for the style change, 25 waiting for cut panels and 15 for a machine failure at the slowest station. That leaves 370 run minutes, in which the line made 400 pieces, and 388 passed first time.
| Measure | Calculation | Result |
|---|---|---|
| Availability | 370 / 450 | 82.2% |
| Performance | 400 x 0.8 / 370 | 86.5% |
| Quality | 388 / 400 | 97.0% |
| OEE | 388 x 0.8 / 450 | 69.0% |
| Balance ceiling | 1.25 / 1.5 | 83.3% |
| Line efficiency | 388 x 20 / (30 x 450) | 57.5% |
Line efficiency alone says 57.5%, and the manager does not know where to go. The split shows that 16.7 points are balance loss for the industrial engineer to fix, and the rest is what happened on the day.
A style change alone cost 40 of the 450 minutes. How to measure changeover time applies well to a line feeding up for a new style.
Traps specific to garments
- Absent operators. With a fixed ideal, a missing operator shows up as slow running in OEE. Line efficiency uses operators present, so it can stay flat while output falls.
- Learning curve. A new style runs slowly for days. Keep the proven rate fixed and name the loss, or the number will reset itself every week.
- Short runs. Small orders mean more style changes, and that caps what any line can score. Compare with your own trend.
- Stitching time as machine utilization. A sewing machine that stitches for a small part of the cycle is not idle in any useful sense. Machine run-time percentages mislead on manual stations.
- Counting at one station. Count at end of line. A count at a mid-line station includes work that hasn't been finished or checked.
How to measure it
Most factories already have hourly output at end of line, so begin there. Add a log of line-level stops with a reason, since those seldom appear in the production report.
For machines, a simple retrofit sensor gives a run signal on cutters and automats. An hour-by-hour board keeps the line's rate visible during the shift.
We found no neutral OEE benchmark for sewing lines, so do not borrow a number from another plant. Start with your own baseline, using how to set OEE targets.
Which number to run the floor on
Keep line efficiency for planning and labor cost, because your SAM data and your customers already use it. Add line OEE as a weekly diagnostic, so stops and balance stop hiding inside one figure.
On cutters, knitting and automats, make OEE the main number. Start with the machine that feeds the sewing floor late most often.
Line-level stops are what most garment reports leave out. Fabrico takes stops and counts live from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line, and you can read OEE per line and shift next to your own line efficiency figure. See how Fabrico tracks stops.
What people ask next
Can OEE be used on a sewing line?
Yes, at line level, with the line's proven rate as the ideal. It works alongside line efficiency instead of replacing it, as garment line efficiency explains.
What is the difference between SAM and ideal cycle time?
SAM is a standard time that includes allowances, and it is the total work content of one garment. The ideal cycle time for OEE is how often the line can release a good garment, based on what it has proven. See ideal run rate.
Is there a standard OEE for the garment industry?
We found no neutral published figure for sewing lines. The general 85% world-class figure traced to Nakajima was not built for manual sewing, so read is 85% OEE realistic first.
How do you calculate line efficiency in a garment factory?
Multiply output pieces by SAM, then divide by operators x minutes worked. The line efficiency calculator does this with your own numbers.
Should I track each operator's OEE?
No. Use individual efficiency against SAM for that, if your site uses it, and keep OEE at machine or line level. Operator buy-in for OEE explains why scoring people backfires.
Fabrico is a light MES for OEE. It takes stops, speed and counts from PLCs or IoT sensors, with add-on AI cameras counting output where a machine has no usable signal, and operators tag the cause at the line.