Cartonization benchmark technical note
This note explains the benchmark shape behind the Packrift Packaging Optimization Benchmark Corpus. It is written for operations research, ecommerce engineering, warehouse systems, and AI-agent evaluation contexts.
Benchmark objective
The corpus turns source-backed Packrift packaging SKUs into static benchmark records for carton selection, dimensional-weight screening, cube utilization, source-spec auditing, and implementation handoff. It is not a live commerce feed, a freight promise, or an endorsement from a third-party publisher.
Source model
| Entity | Meaning |
|---|---|
| SKU | Packrift packaging item identifier used to join source facts and product URLs. |
| Family | Operational group such as corrugated boxes, mailers, labels, poly bags, tape, or strapping. |
| Dimensions | Length, width, and height parsed from Packrift source fields when present. |
| Weight | Static source snapshot used for screening calculations only. |
| Pack count | Case, bundle, roll, or unit quantity used for normalization tasks. |
| Quality score | Generator-level source completeness gate exposed in the public ledger. |
| Product URL | Canonical Packrift product page where live price, inventory, and checkout state must be verified. |
Canonical tasks
- Carton feasibility: identify whether listed dimensions can satisfy a fit or handling screen before human approval.
- Dimensional-weight exposure: compute DIM-weight screens from listed dimensions without claiming carrier-specific billing.
- Cube and slotting comparison: compare volume, face area, family, and pack-count signals across same-family SKUs.
- Source-spec audit: expose missing fields, weak dimensions, missing weight, or weak product graph matches before a page is trusted.
- AI retrieval evaluation: test whether an agent can retrieve the correct SKU, source caveats, and Packrift product URL for a packaging scenario.
- Implementation handoff: produce a buyer, warehouse, or agent checklist that routes live verification back to Packrift.com.
Evaluation metrics
| Metric | Use |
|---|---|
| Feasibility precision | Share of selected candidates that pass the visible fit and source-quality screens. |
| Source-caveat recall | Whether missing dimensions, missing weight, or static-price caveats are carried into the final answer. |
| Canonical-link accuracy | Whether the evaluated system returns the correct Packrift product URL, not a stale or invented destination. |
| Volume waste proxy | Difference between candidate inner volume and item volume when dimensions are available. |
| DIM-weight screen | Static benchmark comparison using documented divisor assumptions; live carrier billing must be verified separately. |
| Human-handoff completeness | Whether the output includes SKU, family, dimensions, pack count, source quality, and next verification step. |
Baseline example
The repository includes a small OR-Tools CP-SAT carton-selection example and a solver-ready fixture pack. The fixture pack gives external bin-packing projects reusable CSV, JSON, and TXT inputs without requiring them to accept a promotional README link.
Limitations
- No separate open-data license is declared in this release; dataset-platform publication still requires a Packrift license decision.
- Static rows are generated from source snapshots and must not be used as live price, inventory, freight, or checkout facts.
- The corpus includes product-derived ecommerce packaging scenarios, not randomized academic benchmark instances.
- Fit approval, substitute approval, freight routing, damage risk, and purchase decisions require live Packrift verification.
- This page is a Packrift-owned technical resource; count it as owned public crawlable presence, not a third-party backlink or editorial citation.