True Shape Nesting With Groups — Why a 95-Part Sheet Isn't One Packing Problem

If you have ever compared nesting tools, you have seen the headline number: parts per sheet. Tool A fits 92, Tool B fits 95, and the one with the bigger number wins. It is a clean, comparable metric — and for a print shop running real orders, it is also the wrong one to optimize alone.

A production sheet is almost never one flat pile of parts. It is a sheet of orders. Order #1041 has 6 keychains. Order #1042 has 3. Order #1043 has 11. They are printed and cut together to save material — but they have to come off the cutter as separate orders, ready to bag and ship. The moment you optimize only for "most parts per sheet," you scatter every order across the whole sheet, and you move the cost you saved on material straight onto the person doing the sorting.

This is what group-aware true shape nesting solves. Below: what true shape nesting is, why grouping makes it a genuinely hard problem, and how Pressria Bridge handles it.

First — what true shape nesting actually is

Rectangular nesting — the packing behind a standard sticker gang sheet — treats every part as the smallest box that contains it. Simple, fast, and wasteful: a keychain shaped like a cat, a star, or a bottle leaves large empty corners inside its bounding box, and those corners are paid-for material you throw away.

True shape nesting places parts by their real outline instead. An irregular part can tuck into the concave gap of its neighbor — the curve of one keychain nestling into the notch of the next. The standard way to compute this is the No-Fit Polygon (NFP): for any two shapes, the NFP describes every position where they touch but do not overlap. Place parts along each other's NFP and you get contact without collision.

The payoff is material. On irregular parts — keychains, acrylic charms, die-cut shapes — true shape nesting routinely recovers 20% or more of the sheet that rectangular packing wastes. That is the part most nesting tools advertise, and Pressria Bridge does it: a hybrid Grid + NFP engine, written in C++, that has placed parts for more than 13,577 production designs across a live keychain pipeline.

But yield is only half of what a production sheet needs.

The part nobody puts on the box: groups

Picture the densest possible sheet. Every gap filled, every concave notch used, 95 parts where a rectangular nester managed 80. Beautiful — until you walk to the cutter table after the cut and try to fulfill order #1042.

Order #1042 is 3 keychains. On a yield-only sheet, those 3 parts went wherever they happened to fit best — one near the top-left, one buried in the middle, one at the bottom edge. To ship #1042, an operator now has to find all 3 among 95 look-alike pieces, confirm nothing is missing, and not accidentally grab a piece from #1041. Repeat for every order on the sheet.

That hunt is invisible in a tool comparison. It does not show up in "parts per sheet." But on a shop floor running 50 to 500 orders a day, sorting a scrambled sheet can cost more operator time than the prepress the nester just automated. The bottleneck didn't disappear — it moved downstream, where nobody measured it.

A production-grade nester has to optimize two things at once:

• Yield — fit as many parts on the sheet as the material allows.
• Group integrity — keep each order's parts together, so the sheet comes off the cutter already sorted.

These two goals pull against each other. Perfect group integrity means reserving a tidy rectangular block per order, which wastes the space between blocks. Perfect yield means ignoring groups entirely. The real job is the trade-off in the middle — and that trade-off is genuinely hard.

Why the boundary between groups is the hard part

Inside a single group, true shape nesting is "just" the NFP problem: pack these parts as tightly as their outlines allow. Solved well enough.

The difficulty lives at the boundary between groups. When group A is packed and group B starts, the parts of B want to interlock with the parts of A — that is what true shape nesting does. But if you let them, A and B interleave, and group integrity is gone. If you forbid it completely, you draw a hard rectangular fence around each group and waste the irregular space along the fence line.

The good answer is neither. It is to let groups share a boundary shape — to let group B's outline nestle against group A's outline the same way two parts do — while still keeping every part's group identity intact and traceable. In effect you nest the groups themselves as true shapes, not just the parts inside them. This is a recursive, two-level version of the NFP problem, and it does not have a clean textbook solution. It is one of the genuinely open problems in 2D nesting, and we have written separately about why we keep obsessing over the space between groups.

How Pressria Bridge handles it

PB's nesting engine runs a multi-stage cascade rather than a single pass. In simplified terms:

1. Prepress reduction — each part's outline is simplified to a clean nesting polygon, so the NFP math is fast and stable.
2. Group classification — parts are tagged by their group (order, customer, or cut job) before placement, so group identity is an input to nesting, not an afterthought.
3. True Shape NFP placement — the primary pass nests parts along their real outlines, group by group.
4. Staircase Y-scan and gap-fill — secondary passes recover the awkward spaces a single NFP pass leaves behind, including the boundary regions between groups.
5. Final grid-scan retry — a last sweep to seat any part that the earlier stages could not place.

The result is a sheet that is dense and grouped: each order's parts stay in a coherent region, the boundaries between orders are packed instead of fenced off, and every part on the output remains tied to its source order. When PB pushes the layout into Illustrator over the live CEP connection, the parts arrive grouped — not as 95 anonymous shapes, but as the orders they came from.

This is why we describe PB as a production line rather than a toolkit. A toolkit hands you a tight nest and leaves the sorting to you. A production line keeps the order intact from the customer's file all the way to the bag.

The comparison that actually matters

When you evaluate a nesting tool, "parts per sheet" is the easy number to compare — and the one to be most skeptical of. Two questions tell you far more about whether a tool was built for production or for demos:

• Does it nest by true shape, or just by rectangle? Rectangle-only packing — grids, honeycombs, MaxRects — is fine for uniform parts, but on irregular shapes it leaves 20%+ of the sheet on the table. If a tool advertises "nesting" but every part still sits in its own box, it is not true shape nesting.
• Does it keep groups intact? A tool that pools every part for maximum density and ignores which order each part belongs to has handed you the sorting problem. For one-off jobs that is fine. For a shop processing many orders per sheet, it is a hidden tax.

Plenty of tools answer yes to the first and no to the second, because the first is visible in a screenshot and the second only hurts on a real shop floor. Group-aware true shape nesting is the combination — and it is the combination because production needs both.

Run your current tool through this. The more boxes you cannot tick, the more sheet space and operator time you are quietly losing:

A tool that ticks the first two but not the rest is a nesting feature. A pipeline that ticks all six is built for a shop floor.

What this isn't

A few honest boundaries:

• The boundary problem is not "solved," it is engineered. Group-boundary packing is an open problem; PB's cascade is a strong practical answer validated in production, not a proof of optimality. We treat it as something to keep improving.
• Group nesting does not beat yield-only on raw density. By design, it gives up a little sheet density to keep orders together. If you genuinely run one order per sheet, that trade-off does not apply to you — use yield-only and ignore this entirely.
• This is about layout, not cutting. PB automates the sheet up to the print-ready Illustrator file. The cut itself runs on your existing cutter and its own software. PB stays maker-agnostic on purpose.

The takeaway

True shape nesting is what saves material. Group-aware nesting is what keeps that saving from leaking back out at the sorting table. A nesting tool that does only the first gives you a great screenshot and a scrambled sheet. A production pipeline does both — fits the parts by their real shape, and keeps every order intact from input file to finished bag.

That is the problem Pressria Bridge was built around, and the 13,577-design pipeline is where it was proven.

---

Pressria Bridge is a Windows desktop application that automates print production workflows including true shape nesting, background removal, cut line generation, and Illustrator integration. Free trial available at pb.pressria.com. Related reading: Sticker Gang Sheet Automation, True Shape Nesting for Acrylic Keychains, NFP Nesting for Print, and What is Nesting?