Fabric to floor — Streetwear Classy

Built by Noorflows (noorflows.com) for Streetwear Classy. Working system, running on demo data. Every order, customer, style and piece count is invented.

Where it starts, and why that matters

Fabric is 60 to 70 per cent of what a garment costs. The standard measure of production waste, the cut-to-ship ratio, begins counting after the fabric has already been cut — which is why a practitioner argument calls it "an incomplete judge". In their worked case the sewing floor showed 4.8 per cent loss, and adding the cutting room revealed roughly 10 per cent overall.

So this starts at the roll. Fabric bought, fabric laid, pieces cut, pieces packed — one balance sheet, and at every step the same two questions: how much is missing, and did anybody write down why.

The fabric

Bought
9300.00 m across 5 lots
Laid on the table
7190.00 m
Still on the shelf
1735.00 m — not cut is not lost
Written off with a reason
239.00 m — roll tails, faults cut around. Ordinary
No record at all
136.00 m

And separately, on one lot every metre is accounted for and the garment still takes 0.60 m more fabric than the costing assumed. Nothing is missing and nobody has done anything wrong. On a long run that is the larger number of the two, and it never appears on a floor report precisely because no metre is unexplained.

Consumption is only reported when it is over the costing by more than three per cent and more than five centimetres. A percentage alone waves through a large garment; a flat floor alone flags every small one.

It stays in metres, never money. What a metre is worth depends on the contract, and a floor system is not where that lives.

What the floor tracker does

It reconciles every cut ticket from the cutting table to the packing bench. A cut ticket says 500 pieces went onto the floor; the packing bench counted 464. Both numbers exist in every system on the market, on two different screens. What none of them puts in front of you is the 36 in between, and the split inside it.

Recorded loss
Somebody wrote it down. Rejects at QC, a bundle scrapped after rework. This is ordinary manufacturing and it is not a problem.
Unaccounted
Nobody wrote anything. The pieces were cut, they are not in a box, and no line on the floor says why. This is the number the build exists to produce.

The two are never added together. Putting them in one total would make a normal week look like a disaster and bury the only figure worth acting on.

The measured week

Pieces cut
3310, across 8 tickets
Reached a box
2920
Still being made
300 — not finished is not the same as missing
Written off with a reason
38 — ordinary
No record at all
64
Packed against a ticket that did not cut them
12 — another order is quietly short

2920 + 38 + 64 + 300 − 12 = 3310, which is the number cut. The sum is printed on the page rather than described, because a page a reader can disprove with mental arithmetic is a page whose other numbers stop counting.

It is deliberately not a "we reduced your losses by this much" figure. We have not reduced anybody's loss and could not measure it if we had.

How the week was built

Seven problems were planted deliberately and seven ordinary situations were planted just as deliberately — a fabric fault written off with a reason, a lot still on the shelf, a garment a few centimetres over plan and inside the allowance, a ticket with rejects written off at QC, a stage that recorded its own rejects, a ticket still on the floor, and a box holding pieces from three tickets that are all the same dye batch. The tracker found 7 of 7 and flagged 0 of 7.

The second group matters more than the first. A floor loses pieces and metres every day; a tracker that shouts about routine rejects and roll tails is one the floor stops reading by Wednesday, and then it is not there for the ones that matter.

The control nothing is wrong with

Fabric is dyed in batches, and two batches of the same colour are not the same colour. Under factory lights nobody can tell; laid side by side in a shop the difference is obvious and the delivery comes back. Every piece involved is sound, which is exactly why a system built around counting defects cannot see it at all.

The tracker reports two cases: one box holding more than one dye batch, and — the case a per-box rule misses entirely — a single customer's order split across batches where every individual box is internally consistent. Those boxes get opened together.

It reports the mix and never judges it. Plenty of buyers accept mixed lots across a bulk order and say so in the contract. That is a commercial decision and it is not ours.

What it cannot tell you

It reads what the floor wrote down
A bundle moved without being scanned is invisible to it, and the answer for that ticket will be wrong in the direction of looking worse than reality.
It cannot say where the unaccounted pieces went
It can say which department they were last counted in, which is the question somebody can go and ask.
It does not know whether a reject was fair
A piece written off with a reason is treated as accounted for, whatever the reason says.
It counts pieces, not money
What a piece is worth depends on the order it was for, and that does not live in the floor system.
It corrects nothing
Every figure goes to a person and the records stay exactly as the floor wrote them.

Demo data throughout. No real production record has been copied into this system, including for testing. The interactive version · How this handles your data