Packaging design

Box dimensions and order-level packing were evaluated together, allowing the packing result to guide the next candidate set of box sizes.

Industry: Logistics / Packing
Decision family: Spatial planning
Evaluation scope: PoC test-data evaluation

Customer identity and identifying operational details are withheld.

The operating decision

A box-size set must reduce empty space while still allowing the items in every evaluated order to be packed.

Inputs: Item types and quantities per order · Item dimensions · Candidate box dimensions
Decision: Set of box sizes · Box selected per order · Item orientation and 3D placement
Constraints: Rotation rules · No overlap · Box boundaries · Packability of each evaluated order

How it is approached

Check packing order by order
Place and rotate the actual order items, then check overlap and box boundaries rather than relying on total volume alone.
Select among packable boxes
Try each order in candidate boxes and choose a suitable size among the candidates that can accommodate it.
Feed packing results back into size design
Aggregate fill and packability results, propose a new box-size set and repeat the evaluation.

Box-size search informed by order-level packing.

Pause animation

INPUT

Customer order data

As the order mix changes, empty space and packing failures appear

Item types and quantities per order

Length, width and height per item

→Orders in

Packing evaluation per order

Pack it for real, then choose the box

1 · 3D packing checkrotation, overlap and boundary checks

2 · Box choicea size that packs is chosen

3 · Results talliedfill rate and share of packable orders

→Score

Box size design

Propose the next size set

N candidate box sizeslength × width × height combined per box

The packing score feeds the next candidates, and the loop repeats.

Next box-size candidates passed on

A packaging design structure: 3D packing checks score each candidate box set, then feed the next search. The animated boxes illustrate candidate changes, not a live solver.Scroll to explore the diagram ↔

Original side-by-side order-level packing results. — Original order-level packing comparison from the project’s test-data evaluation.

PoC test-data evaluation

Validation & scope

The source reports a PoC comparison on test orders. Its working examples show the method; calculation-time comparisons come from a separate final review. The evaluation supports box-size selection and does not establish operating-cost savings.