Technical
What Is 3D Bin Packing? How Cartonization Places Items in Boxes
3D bin packing is the optimization problem of placing rectangular prism items into a rectangular prism container (a box). Academically it is NP-hard; production systems use heuristic engines for speed and determinism.
Problem definition
Input: container dimensions, item dimensions (each with quantity), legal rotations, weight limit. Output: (x, y, z) coordinates and rotation per item, or leftovers that cannot be placed.
Shipping cartonization is bin packing enriched with business rules, multiple boxes, and a rule engine.
Production requirements
Before marketing says “AI packing,” the engine must do the following:
- Legal orientation (which axes can rotate)
- Bounding-box collision checking
- Container bounds (no face extends outside)
- Box weight limit
- Multiple boxes and leftovers
- Deterministic scoring
Why volume math is not enough
∑(item volume) ≤ box volume is not sufficient in the real world. Items cannot always stack, voids form L-shapes, and a long edge can exceed a carton dimension.
Volume-only solutions constantly produce overrides at the pack bench.
WiCartonize engine
It uses extreme-point / free-space placement, multiple item-ordering strategies, and box-selection scoring. Results are visualized in Inspector with millimeter coordinates and rotation.
See engine and features for product capabilities.
Optimal solution or heuristic?
Optimal solutions are impractical for real-time fulfillment on most orders; consistent heuristics are the production standard.
How are soft goods (apparel) modeled?
Compressibility rules or separate profiles; the engine starts from rigid prism assumptions.