How much cutting-table time the order of a cut file costs

A cutting table cuts a file in the order the file is written, unless its own software re-orders it. That order is usually an accident of how the drawing was made: shape by shape, layer by layer, or in whatever order a CAD export walks its records. The knife pays for it in tool-up travel, tool changes and tool-downs, on every sheet of the run.

AutomatePrint now has a free tool that shows how much. Cut sequence, in Equipment → Tools & dies, loads a CF2, DXF, PDF or Zünd Cut Center (ZCC) cut file, plays your cutting table through it as written and in an optimised order, compares the time per sheet and per job on your own cutter, and exports the faster file. It is free on every plan, including Free: in the app, or directly through the AutomatePrint API with your API key.

Where the time goes

Cutting time is more than cut length divided by speed. On a flatbed with a tangential knife, every sharp corner is a lift, a turn and a plunge. Every separate path costs a plunge and a lift-off. Between paths the head travels with the tool up, accelerating and braking. Switching from the creasing wheel to the knife and back costs a tool change each time. The cut length is fixed by the design, but travel, tool changes and the number of tool-downs depend only on the order. So does something that costs more than time: a part whose outline is cut before the holes inside it is already loose when the knife comes back, and can shift.

What the optimiser does

Choosing the order is a version of the travelling salesman problem with extra choices and rules: each open line can be cut from either end, each closed shape can start at any point, every tool's lines go together, and everything inside an outline must be cut before the outline is finished. The optimiser works in steps a production planner would recognise:

• Tools in order. Creases and scores while the sheet is whole, then perforations and kiss cuts, thru-cuts last, so the head changes tool once per tool.
• Parts never move. Holes, windows, slits and nested parts come before the outline around them.
• Fewer tool-downs. Lines are split where they meet and joined into long runs. Rule drawn twice, such as the shared edge of two shapes drawn side by side, is cut once (an option you can turn off).
• Least travel. A nearest-neighbour start, then the best start point and direction for every path in that order (solved exactly), then 2-opt and Or-opt moves that reverse or move runs of paths, each checked against the rules, until nothing more is saved. Travel is weighed in seconds of acceleration and braking, not millimetres.

It is deterministic, so the same file and settings always give the same order, and a die with a few thousand paths takes a second or two.

What it saved on sample dies

We ran four generic sample dies through it: cartons from the box designer's reverse tuck end, written by the app's own die export (one shape per station, creases and cuts per carton), and plain labels and butt-cut tags. No customer files. The times come from the cutting-time model the app uses for estimates, with each machine's published specifications, the app's estimated defaults for corner lifts, plunges and overcut, and a 2-second tool change. They are simulated, not stopwatch timings on a machine, and they leave out setup, sheet handling and registration, which are the same in both orders. The file's order is taken the way most cutter software runs it: each tool's lines together (all creases, then all cuts), in the order the file lists them, so both sides change tool once.

The roll-cutter rows show how a different machine changes the result: a roll cutter with a drag knife never lifts at corners, so travel is a larger share of its time. It would not crease board in practice.

On the 24-up carton the knife travels 16.7 m instead of 60.7 m per sheet. As written, 120 of its cuts come after the outline around them is finished, so those parts are already loose when the knife comes back. Over 250 sheets the optimised file saves about 4.4 hours of table time on the flatbed.

Two of those numbers need context. The tags file draws every tag's four edges separately, so shared knives are cut twice and the knife goes down 118 times; joined into straight knife lines it goes down 19 times. A CAD file that already draws long knife lines saves far less. And a cutter that follows the file line by line, without sorting by tool, would switch between the creasing wheel and the knife 47 times on the 24-up carton; against that order the optimised file saves 18 % on the flatbed and 31 % on the roll cutter.

Try it on your own files

The saving depends on how your files are written and on your cutter, so measure it rather than trust ours. Open Equipment → Tools & dies → Cut sequence, drop a cut file, choose your cutter and press play. The job calculator turns the per-sheet difference into hours and cost at your hourly rate, and the optimised file downloads as PDF, DXF, CF2 or ZCC with its line types kept. A ZCC job can keep its own curves, bridges and register marks and only change its order, with the material, copies and job name set for the cutter. To build it into your own workflow, the same steps are four API calls: upload, analyse, export and save. The how-to guide walks through both, with the requests and responses. Imposition and estimating cut files from AutomatePrint are written in the optimised order already.

One caution: if your cutter's software re-optimises the order on import, it cuts its own order, not the file's. Check which your cutter does before you compare. Zünd Cut Center, for one, sorts a job's outlines itself unless the job marks its methods as custom-ordered; the ZCC files this tool writes do, so Cut Center cuts them in the optimised order.