The Finishing Room: Iron, Fold, Pack Without Losing Pieces

Every factory I have visited treats the finishing room as the place where the work is already done. The sewing is finished, the hard part is over, what remains is pressing and folding. Then at lot close the count is short by two or three hundred pieces and nobody can say where they went.

I run a CMT factory in Nepal and this was our own blind spot for a long time. The sewing floor was fully tracked — every bundle scanned at every operation — and the dispatch side was documented, because buyers demand documents. In between sat a room where garments came in on trolleys and left in cartons, and the only record was what the packing supervisor remembered.

What the Finishing Room Actually Does

Most CMT factories run roughly the same sequence, with local variations in order and in who does what. The names differ — finishing, packing section, post-sewing — but the work is consistent.

1. Receive from sewing. Garments arrive from the last sewing operation, usually in bundles or trolleys, usually grouped by lot, article, colour and size.
2. Thread trimming. Loose ends and chain stitches from the last operations are cut. In smaller factories this happens at the sewing table; in larger ones it is a dedicated station.
3. Checking. A checker examines the whole garment rather than a single operation: measurements, symmetry, stitching, stains, holes, missing labels. This is the last chance to catch anything before it becomes a buyer complaint.
4. Spot cleaning and repair. Stains go to the cleaning table, defects go back for rework, and pieces that cannot be brought to A-grade are downgraded.
5. Pressing. Iron or steam press, depending on the fabric and the buyer specification.
6. Folding and tagging. Fold to the buyer's specification, insert tissue or cardboard if required, attach hangtags and price tickets, then polybag.
7. Packing. Pieces go into cartons by the buyer's packing method, cartons get marked and weighed, and the lot becomes a shipment.

Why Pieces Go Missing Between Sewing and Packing

When a lot closes short, the pieces are rarely stolen and almost never destroyed. They are misfiled. Here is where they actually go, in rough order of how often I see it.

• Sitting in the repair pile. A piece went back for rework three days ago and is still in the basket under the table. It was counted as sewn, it has not been packed, and nothing in the system says it exists.
• Downgraded quietly. A checker pulls out a stained piece, puts it on the B-grade shelf, and tells nobody. The A-grade count drops with no corresponding record.
• Mixed into the wrong lot. Two lots of the same article in different colours are being finished at the same time. A trolley gets combined, and a hundred pieces of lot A ship as lot B. Both counts are wrong and they cancel out only if you are lucky.
• Packed but not recorded. The carton is genuinely built and genuinely correct, but the packing record was made on a sheet that went missing before the challan was raised.
• Sample pulls. Merchandising takes six pieces for the buyer, QC takes four for the inspection sample, and nobody writes it down. This one is small but it is constant.
• Never sewn at all. Sometimes the shortfall did not happen in finishing. Cutting issued fewer pieces than the sheet says, or a bundle was lost on the sewing floor, and finishing takes the blame because that is where the shortfall is first visible.

The Finishing Receipt as a Control Point

The change that fixed this for us was not a new department or more staff. It was one scan at the boundary.

When goods are formally received from the sewing floor into finished-goods inventory, someone scans and the system writes a receipt: which lot, which article, which colour and size, how many A-grade, how many B-grade. Our receipts are numbered atomically in the format SAC-BSYEAR-NNNN and generated inside a database transaction, so two people receiving at the same time cannot produce the same number or skip one. A gap in the sequence would be a real event, not a numbering artefact, which is exactly the property you want in a document a buyer might audit.

The scan gives immediate physical feedback: a large green or red confirmation on screen, and an automatic announcement over the factory PA. That sounds like a small detail. It is not. The quiet failure mode in every scanning system is the scan that does not register while everyone assumes it did, and a confirmation you can hear across the room removes it. The same principle runs through our whole floor, and I have written about it in more detail in the piece on end-to-end tracking without data entry.

Once receipts exist, the comparison that was previously impossible becomes routine: sewn output for the lot against received output for the lot, checked the same day. If eighty pieces of the yellow medium were reported sewn and sixty-four were received, you go and look this afternoon, while the trolley is still in the room, rather than at lot close six weeks later when the trail is cold.

Checking, Repair, B-Grade, Pack: The Fork in the Middle

A piece that fails checking has three possible destinations, and the finishing room runs well or badly depending on how deliberately that choice is made.

The first is repair. In our system the damaged pieces are split out into repair bundles with their own QR codes, so the rework is tracked as work rather than vanishing into a basket. Repair pays 1.5 times the operation rate, because nobody chooses to redo someone else's seam at the normal rate and pretending otherwise just means the pile grows. The detail of that flow is in the article on spot cleaning and rework.

The second is B-grade. We track B-grade inventory by defect type — stain, tear or hole, size issue, stitching defect — because the destinations differ. Stains sometimes recover with a second cleaning attempt; a hole does not recover at all; a size issue is a perfectly well-made garment that is simply the wrong size and may sell as a second at full construction quality. One undifferentiated B-grade pile means re-sorting by hand the day someone offers to buy it.

Whichever way a piece goes, the count has to follow it. The reason short shipments are so common is that the fork is real but the counting treats the room as a single pipe. I have gone through that arithmetic properly in the article on short shipments and dispatch accuracy.

Measuring Finishing Throughput

Sewing has a well-developed measurement vocabulary. Finishing usually has none, which is why it is the first place a production manager guesses. A few things are worth measuring, and they are all available from receipts and checking records without any new paperwork.

• Pieces received per day, by lot. The basic output number. Plot it against sewn output and the gap between the two lines is your work in progress inside the room.
• A-grade percentage at receipt. How much of what arrives passes straight through. A sudden dip usually points at one operation on the sewing floor, not at finishing.
• B-grade split by defect type. Which category is growing tells you where to look. A rising stain count is a handling and housekeeping problem; a rising size-issue count is a cutting or pattern problem.
• Age of the repair queue. Not the size of it — the age. A repair pile of two hundred pieces that clears daily is healthy. A pile of forty pieces where the oldest is three weeks old is a lot that will close short.
• Pieces stuck between receipt and packing. If receipts say four thousand and dispatch says three thousand two hundred, eight hundred pieces are physically in the room. Somebody should be able to point at them.

Where Piece-Level Tracking Honestly Ends

I want to be straight about a boundary that most software marketing blurs. Through cutting and sewing, the bundle QR carries lot, article, colour, size, component and quantity, and every scan records the operator, machine, operation and timestamp. That is genuine piece-level attribution.

It ends at folding. Once a garment is pressed, folded and polybagged, it has no scannable identity of its own unless you are printing and applying a unique code to every piece, which almost no CMT factory does and no buyer in our segment pays for. From packing onwards, the honest unit of tracking is the lot and the article. Our receipts, our challans and our pipeline balances all work at that level.

Common Failure Modes

• Finishing treated as a buffer, not a process. When the sewing floor is behind, garments are pushed into finishing to make the sewing number look good. The room fills up, the queue stops being visible, and losses start.
• Checking done by the same person who packs. Under pressure to ship, the checker becomes lenient, because the person who has to explain the delay is the same person deciding what passes. Separate the two roles even in a small factory.
• No record of the sample pulls. Merchandising, QC, and the buyer's visiting inspector all take pieces. Six here, four there, twelve for a photo shoot. Keep a book.
• Two lots in the room at once, no physical separation. This is the single most common cause of mixed shipments in my experience. Tape on the floor is cheap.
• Repairs with no rate. If rework is unpaid or paid at the normal rate, it queues. It is a workload question, not an attitude question.
• Packing to the wrong ratio. Perfectly good garments packed against the previous season's size ratio. The arithmetic and how to check it are in the article on packing methods.

Where to Start

If your finishing room is currently untracked, do not start by installing a system. Start by drawing the boundary and counting across it for a week: what came in from sewing, what went out to packing, what sits in repair, what sits in B-grade. Do it on paper if you have to. Almost every factory that does this for a week discovers the gap is larger than they assumed and is concentrated in one or two places.

Frequently Asked Questions

What is the finishing process in a garment factory?

Finishing is everything between the last sewing operation and the packed carton: thread trimming, final checking of the whole garment, spot cleaning and repair of what fails, pressing, folding, tagging and polybagging, then packing into cartons. Some factories also run a final measurement audit inside the room. It is a quality process as much as a handling process, because the checking step in the middle is the last point at which a defect can be caught before it becomes a buyer complaint.

Why do garment factories lose pieces between sewing and packing?

Because most factories record the last sewing scan and then count cartons at dispatch, with nothing recorded in between. Pieces sit in the repair basket, get quietly moved to the B-grade shelf, mix into a neighbouring lot of the same article, or are pulled as samples with no note. None of these is theft and all of them produce a shortfall at lot close. Recording a formal receipt at the boundary of the finishing room makes the loss visible the same day instead of at the end of the lot.

What is a finishing receipt and why does it matter?

It is the record created when finished goods are formally received from the sewing floor into inventory: lot, article, colour, size, quantity, and how many pieces are A-grade against B-grade. In our system these receipts are numbered atomically in the format SAC-BSYEAR-NNNN inside a database transaction, so numbers cannot be duplicated or skipped, and each scan shows a large green or red confirmation on screen and announces over the factory PA. It matters because it gives you a number to compare against sewn output while the goods are still in the room.

Can you track individual garments through the finishing room?

Up to a point. The bundle QR carries lot, article, colour, size, component and quantity, and every scan through cutting and sewing records the operator, machine, operation and timestamp. That piece-level identity ends at folding: once a garment is pressed, folded and bagged it has no scannable identity of its own unless you apply a unique label to every piece, which most CMT factories do not. From packing onward the honest unit of tracking is the lot and the article, and any vendor claiming otherwise should be asked exactly which label they are printing.

How do you measure finishing department performance?

Use numbers the receipts already produce rather than inventing new paperwork: pieces received per day by lot, the A-grade percentage at receipt, the B-grade split by defect type, the age of the oldest item in the repair queue, and the gap between total received and total dispatched. The age of the repair queue is the most underrated of these, because a small pile of very old pieces predicts a short shipment far better than a large pile that clears every day.

The finishing room is not a difficult process. It is a process that most factories have simply never written down, which is why it is measured by memory and reconciled by argument. One scan at the boundary turns it into something you can look at on a screen at eleven in the morning, which is early enough to do something about it.

Santosh Rijal is the founder of Scan ERP, a garment manufacturing ERP system designed for factory floor operations. He works directly with sewing lines, cutting rooms, and production supervisors across Nepal's garment manufacturing sector.