Javascript must be enabled for the correct page display

Analysis of new models of the three-dimensional bin packing problem for the application to the parcel delivery service industry

Kho, Margot (2018) Analysis of new models of the three-dimensional bin packing problem for the application to the parcel delivery service industry. Research Project, Industrial Engineering and Management.

[img]
Preview
Text
Research_Project_M_Kho_27sept2018.pdf

Download (1MB) | Preview
[img] Text
toestemming.pdf
Restricted to Registered users only

Download (210kB)

Abstract

The unceasing increase of e-commerce volumes worldwide has been putting under pressure the sector of parcel delivery service. Due to a highly competitive market, the different logistic service providers strive to be efficient in every segment of the logistic chain. Among these segments, the packing of parcels into containers is a crucial aspect that, if not done properly, may cause additional costs, such as the use of more resources than the ones actually needed. In this research, we investigate such a packing problem and model it as a three-dimensional bin packing problem (3D-BPP). The 3D-BPP can be defined as the placement of a set of different parallelepiped sized items parallel to the bin dimensions that have to be placed in a minimized number of bins. This problem is a well-known NP-hard problem. We start by discussing the model proposed by Chen et al. (1995), where items may rotate 360◦. Next, we develop mathematical models with flexible item rotations and the possibility to include layers, i.e. shelves, by adding constraints to the model of Chen et al. (1995). Even though the new models solve different problems, still their solutions may be used as an indication for the basic model. Our goal is to test the computational performances of an industrial solver, performing a branch-and-bound procedure, for several input classes among which also instances generated using real-world data from a Dutch parcel delivery service provider. Results show that for larger instances we can obtain better results mainly regarding the computational time, with some of the newly developed models. This can potentially open opportunities for enhanced automation and improvement in the packing process.

Item Type: Thesis (Research Project)
Supervisor name: Fazi, S. and Sierksma, G.
Degree programme: Industrial Engineering and Management
Thesis type: Research Project
Language: English
Date Deposited: 27 Sep 2018
Last Modified: 05 Oct 2018 14:34
URI: https://fse.studenttheses.ub.rug.nl/id/eprint/18644

Actions (login required)

View Item View Item