Dynamic Order Acceptance and Scheduling in On-demand Additive Manufacturing Using Powder Bed Fusion Technologies: A Strategy-based Heuristic Approach

Abstract

The emergence of Additive Manufacturing (AM) based on-demand manufacturing has presented new decision-making challenges in order processing that require innovative and practical solutions. The study presented in this thesis is motivated by this exigency and aims to propose a novel strategy-based heuristic approach to effectively address the Dynamic Order Acceptance and Scheduling (DOAS) problem within the context of platform-based on-demand manufacturing using Powder Bed Fusion (PBF) technologies. Although operations research in AM has garnered significant attention in recent years, with numerous exact, heuristic and metaheuristic algorithms being proposed to address nesting and scheduling problems, the DOAS problem has not received much attention due to its highly complex nature. Therefore, this thesis comprehensively presents the analysis of the critical challenges associated with the DOAS problem and the factors contributing to these challenges to provide a clear introduction to the problem. A mathematical model is developed to formally define the problem and to facilitate the optimization of the DOAS decision-making process by parameterizing the constraints and decision variables. Furthermore, novel strategy-based heuristic algorithms are proposed for optimizing the decision-making process in DOAS. Especially, two interrelated heuristic decision-making procedures are implemented for generating feasible AM jobs on individual PBF machines and approving them a system perspective. In this context, local and global decision-making strategies are designed to aid in the selection of candidate part orders by PBF machines and to facilitate decision-makers in selecting feasible AM jobs for approval, respectively. A comprehensive computational study is conducted to validate the developed heuristic DOAS algorithms and evaluate the performance of the proposed decision-making strategies. The experimental results indicate that the implemented heuristic algorithms can effectively solve the DOAS problem and relative optimal outcomes can be obtained by employing one of the six identified sets of decision-making strategies. Additionally, the combined applications of identified decision-making strategies are demonstrated to investigate the effects of changes in market demands, variations in promised due dates, and idle costs of PBF machines, providing valuable insights into employing decision-making strategies for specific scenarios.