Three-dimensional packing problem (Real-world datasets)

Published: 12 February 2025| Version 1 | DOI: 10.17632/kxcb2frrnv.1
Contributors:
Sara Ali,
,

Description

Introduction The dataset comprises 198 real instances obtained from a company, each instance detailing a specific packing scenario. These scenarios involve packing a set of 3D items ranging from weakly heterogeneous to strongly heterogeneous into the minimum number of identical 3D containers. Dataset Structure Each instance is stored in a separate .txt file, and provides a unique configuration of items and containers, serving as an individual test case for packing algorithms. The files are organized as follows: Line 1: Contains three integers representing the dimensions of the container: Length, width, and height respectively, Line 2: Contains a single integer indicating the number of items to be packed, Line 3 and subsequent lines: Each line represents an item and contains four fields: Item id, length, width, and height respectively. File Naming Convention Each file is named according to the specific instance it represents. The naming convention is as follows: loja_xxxx.txt, where xxxx is a unique instance number ranging from 0000 to 9967. The instance numbers are not sequential and do not follow a specific order.

Files

Steps to reproduce

This dataset can be used for testing and comparing various 3D-PP algorithms.

Institutions

Instituto Politecnico do Porto Instituto Superior de Engenharia do Porto, Universidade do Porto

Categories

Packing Problem

Licence