SyntaxFlow

Glass · Production

Glass offcut inventory: cut from the rack before you buy a sheet

Every cutting optimiser you can find online nests onto virgin sheets. No glass shop works that way. There is a rack against the wall holding last month's offcuts — glass that is already bought, already paid for, and already taking up space. The money in optimisation is not in shaving a point off the yield figure. It is in cutting Tuesday's order out of the rack.

The enquiry on the phone
Your rack
Add an offcut to see The rack is shared with the cutting optimiser — whatever you put here is there too, on this machine only.

What it is worth, on one real job

Forty pieces across four sizes — 12 at 1050 × 850, 10 at 600 × 900, 14 at 450 × 600 and 4 at 1500 × 700 — nested onto 3210 × 2250 stock with 10 mm edge trim and 4 mm breakout. That is the worked example on the optimiser page, so you can reproduce every number below rather than take them from us.

The rack holdsSheets boughtWhat happens
nothing 4 The ordinary answer. 83.4 % yield, 4.1 % scrap — already a good nest.
two whole offcuts
3190 × 2230
2 Half the glass comes off the rack. At 28 €/m² that is 398 € not spent on a single order, and the yield goes up, to 84.0 %.
the same two, plus
a 600 × 500 scrap
2 The small one is not touched. Opening it would not remove a sheet, so it stays on the wall — same 2 sheets, same 84.0 % yield.
only small offcuts 4 The rack is left alone entirely and the job is cut from fresh glass, because nothing on the rack saves a sheet.

That last row is the one that matters, and it is the one most people would not think to build.

The rule: an offcut has to earn being opened

The naive version of this feature is easy and quietly harmful. Pack onto the rack first, take whatever fits, report a lovely number. We built exactly that, tested it on the job above, and it consumed three good offcuts and still bought the same four sheets.

Read that again, because it is the trap. The shop would have lost three labelled, rackable rectangles — turned into smaller tails, harder to find a use for, more pieces to handle at the table — and saved nothing at all. A tool that does that is worse than no tool, because it does it invisibly and calls it optimisation.

So the rule is: a rack piece is only opened when opening it means buying fewer sheets. Everything else stays on the wall. When the rack cannot help, the page says so in plain words rather than quietly cutting into it anyway.

Mechanically, the optimiser solves the job three ways — with no rack, with the whole rack, and then with each opened piece taken away in turn to see whether the sheet count rises. A piece whose removal changes nothing was not earning its place, and it is put back. On a long list with a big rack, where testing every piece one at a time would make the page stutter, it falls back to a cheaper test: any offcut that came back less than half used was opened for too little, and is dropped.

Two details that are wrong in most tools

An offcut takes no edge trim. Edge trim exists because a delivered sheet has a damaged rim. An offcut was cut in your own shop — its edges are already clean, and the size written on the label is usable glass. Trimming it a second time shrinks the rack on paper by 20 mm a side and makes the whole feature look weaker than it is.

Glass is not interchangeable. A 6 mm clear offcut cannot fill a 4 mm low-e order, and a rack that hands you the wrong glass is worse than no rack. So each piece carries what it is, in your own words — "6 clear", "4 low-e", "33.1 lam" — and only matching glass is offered for a job.

Where the rack lives, said plainly

In your browser, on your machine. Nothing about your stock or your customers is uploaded, because there is no server here to upload it to — which is also the honest limit of it: it does not follow you to another computer by itself.

So the rack saves out as a spreadsheet. Keep that file on your own drive, open it in Excel, edit it there if you prefer, and load it on whichever machine is next to the cutting table. It reads a list that is simply width and height in the first two columns, so an inventory you already keep will very likely load as it is.

What comes back on the rack

Cutting a job creates offcuts as well as consuming them, and three kinds of rectangle survive a three-stage guillotine nest: the band above the last strip, the end of a strip past the last column, and the space above the last piece in a column. All three are real, rackable glass. The cutting plan lists every one with its size and its position on the sheet, so it can go back on the wall under a label instead of against it unmarked.

Anything narrower than 150 mm is treated as scrap rather than stock, because counting slivers as inventory flatters the waste figure. That threshold is yours to move.

Try it on your own rack

The cutting optimiser is free, needs no account, and runs entirely in the browser. Paste a cutting list straight out of a spreadsheet, type in the offcuts standing against your wall, and see what it does to the number of sheets you were about to order.

What this is. A production planning aid, not a stock control system and not a signed calculation. It will tell you what a nest costs and what it leaves; it will not tell you whether that piece of glass is really still on the rack. Check the wall before you cut.