Shipping configuration optimization with topology-based guided local search for irregular shaped shipments

Xinyue Chang, Purdue University

Abstract

Manufacturer that uses containers to ship products always works to optimize the space inside the containers. Container loading problems (CLP) are widely encountered in forms of raw material flow and handling, product shipments, warehouse management, facility floor planning, as well as strip-packing nesting problems. Investigations and research conducted two decades ago were logistic orientated, on the basis of the empirical approaches. Starting from the late 1990s, researchers and experts in disciplines, such as mathematics, computer science, and industrial engineering, actively participated in the developments of solutions to CLP. Their contributions are mainly in the areas of typological analyses, heuristic methods, and optimization. However, even in the mid-2010s, the gaps between those research contributions and the applicable level of problem-solving methods for industry are still tremendously huge. Especially, the majority of existing theoretical solutions fall short in the real-world applications due to the incapability of handling irregular shapes in three-dimensional space. To tackle this shortcoming, this research presents a topology-based metaheuristic approach that simplifies both the irregular shapes and the optimization process. Real-world data and constraints are applied to test the validity of this metaheuristic approach, such as compatibility between commodities, orientation limitations, etc. This research is a project of Center for Technology Development (CTD), sponsored by American Axle Manufacturing, Deere & Company, Eaton, and Faurecia. Within the work scope of this research, two distinctive CLP scenarios are constructed. One scenario is the Shipping Configuration Optimization (SCO), targeting the containerization processes of shipping raw materials and irregular shaped Semi-Knockdown (SKD) modules. The generalized constraints of this scenario are custom and highway regulations, availability of the facilities, labor cost, easiness of loading and unloading, etc. The other scenario is Warehouse and Facility Optimization (WFO). The goal of this scenario was to enhance the efficiency of shelf usage, reduce the floor occupation, corresponding to the change of demands, production plans, regulations, et cetera. The developed metaheuristic approach utilizes topological spatial optimization to handle the complexities related to the three-dimensional irregular shapes. Those irregular shaped items are pixelated into three-dimensional envelop shapes, in which some clearances are added onto the original shapes to ensure the final plan is free of interference. In order to deal with the huge amount of combinations of shipping configurations, this approach applies topology-based category optimization and path optimizations to the existing guided local search (GLS). Topology-based category optimization and path optimization significantly reduce the total number of the possible configuration combinations that need to be analyzed, evaluated, and validated. To test the metaheuristic approach, a proof-of-concept end-user program for engineers that implement the heuristic method is designed and developed. Pilot tests using real world data and conditions from the industrial members of CTD are conducted. The proposed heuristic method satisfactorily takes the constraints into account and generated optimal shipping configurations which cannot be obtained from previous research.

Degree

Ph.D.

Advisors

Zhang, Purdue University.

Subject Area

Industrial engineering|Mechanical engineering|Packaging|Operations research

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS