Gang sheet nesting: how automatic packing cuts film waste
Nesting is the automatic arrangement of multiple designs onto a gang sheet so they consume the least possible film. It replaces manual layout work in Photoshop or a drag-and-drop builder with an algorithm that packs the queue in seconds — and, done right, it beats hand-packing on both speed and material use.
The problem, formally
A DTF roll has a fixed width (say 22″) and effectively unlimited length. Each pending design is a rectangle. The goal is to place every rectangle — optionally rotated 90° — without overlaps, minimizing the total length of film consumed. This is strip packing, a classic optimization problem; finding the perfect layout is computationally infeasible at scale, but good heuristics get very close, very fast.
How skyline packing works
A proven approach is the skyline bottom-left heuristic. The algorithm tracks the "skyline" — the current top edge of everything placed so far — and drops each new piece into the lowest, leftmost position where it fits. Sorting pieces tallest-first keeps the skyline flat, which leaves fewer unusable gaps. For each piece, the packer tries both orientations and keeps whichever wastes less trapped area beneath it.
Why determinism matters
A subtle but critical production requirement: the same input should always produce the same sheet. If a sheet jams in the printer or a batch of transfers is lost, the shop needs to reprint that exact sheet — same layout, same labels, same everything. Randomized packers can't promise that. Deterministic ones can: same units in, byte-for-byte identical sheet out.
What good utilization looks like
- Hand-packed sheets typically land between 60–80% utilization depending on patience and skill
- Algorithmic nests on a mixed queue commonly reach 85–92%
- Utilization should be reported per sheet — it is the number that translates directly to film cost
- Per-design labels (QR blocks) must be part of the packed geometry, or labels end up overlapping art
KOVA's engine is deterministic skyline strip-packing with 90° rotation and best-fit candidate selection. Every design gets a QR label reserved beside it as first-class packed geometry, utilization is reported per sheet, and oversized pieces are rejected with a human-readable reason instead of silently dropped.
See Production & NestingFrequently asked questions
For mixed queues at production volume, yes — algorithms consistently hit 85%+ utilization in seconds. A patient human can match the density on a small sheet, but not at 40 designs across 6 sheets, hourly, all day.
Turning a design 90° often lets it fill a gap its original orientation can't. Since DTF transfers have no grain direction, rotation is free space savings — the pressed result is identical.
It physically cannot print, even rotated. Good nesting software rejects it explicitly with the reason, so the operator can split or resize the design instead of discovering the problem at the printer.