Search

The ability to handle unpredictable dynamic events is becoming more important in pursuing agile and flexible production scheduling. At the same time, the cyber-physical convergence in production system creates massive amounts of industrial data that needs to be mined and analysed in real-time. To facilitate such real-time control, this research proposes a hierarchical and distributed architecture to solve the dynamic flexible job shop scheduling problem. Double Deep Q-Network algorithm is used to train the scheduling agents, to capture the relationship between production information and scheduling objectives, and make real-time scheduling decisions for a flexible job shop with constant job arrivals. Specialised state and action representations are proposed to handle the variable specification of the problem in dynamic scheduling. Additionally, a surrogate reward-shaping technique to improve learning efficiency and scheduling effectiveness is developed. A simulation study is carried out to validate the performance of the proposed approach under different scenarios. Numerical results show that not only does the proposed approach deliver superior performance as compared to existing scheduling strategies, its advantages persist even if the manufacturing system configuration changes.

The ability to handle unpredictable dynamic events is becoming more important in pursuing agile and flexible production scheduling. At the same time, the cyber-physical convergence in production system creates massive amounts of industrial data that needs to be mined and analysed in real-time. To facilitate such real-time control, this research proposes a hierarchical and distributed architecture to solve the dynamic flexible job shop scheduling problem. Double Deep Q-Network algorithm is used to train the scheduling agents, to capture the relationship between production information and scheduling objectives, and make real-time scheduling decisions for a flexible job shop with constant job arrivals. Specialised state and action representations are proposed to handle the variable specification of the problem in dynamic scheduling. Additionally, a surrogate reward-shaping technique to improve learning efficiency and scheduling effectiveness is developed. A simulation study is carried out to validate the performance of the proposed approach under different scenarios. Numerical results show that not only does the proposed approach deliver superior performance as compared to existing scheduling strategies, its advantages persist even if the manufacturing system configuration changes.

i ACKNOWLEDGMENTS ii TABLE OF CONTENTS iv LIST OF FIGURES viii LIST OF TABLES x CHAPTER 1: INTRODUCTION 1 1.1. Reverse logistics definitions and characteristics 3 1.2. Designing reverse logistics network for product recovery 7 1.3. Research objectives 9 1.4. Research scope and methodology 10 1.5. Organising the thesis 11 CHAPTER 2: LITERATURE REVIEW 13 2.1. Product recovery 13 2.1.1. Drivers 13 2.1.2. Definitions 15 2.1.3. Remanufacturing 16 2.1.4. Product recovery activities 17 2.1.5. Implementation in industries 19 2.1.5.1. Computer industry 19 2.1.5.2. Single-use camera industry 21 2.1.5.3. Photocopier industry 23 2.2. Reverse logistics: issues and aspects 25 ATTENTION: The Singapore Copyright Act applies to the use of this document. Nanyang Technological University Library

Study the reverse logistics network designs for computer recovery under stochastic conditions of demand for the new products, product returns, and product recoverability by performing simulation using the simulation tool ARENA 7.01. Master of Science (Logistics)

This research aims to develop daily schedules by solving a model that addresses both the batching of raw materials at the RMC plants and the delivery f concrete by the trucks to various customer site for an RMC company. Master of Engineering (MPE)

The objective of this research is to develop a coordination approach for alignment of inventory decisions in decentralized supply chains. First of all, the value of supply chain coordination is investigated and justified in a simple two-echelon inventory model. Based on that, this research develops a supply chain coordination framework, named ASCEND. Doctor of Philosophy (MPE)

This dissertation looks at the author*s organization's current vendor management process and seeks ways to improve it. The focus of this dissertation is to identify the critical areas for improvement and propose solutions to address these gaps. The project was carried out by various means such as use of case studies of external agencies, interviews, implementation of a trial, and a survey to collate feedback for further refinement of the solutions. Master of Science (Logistics)

This dissertation concentrates on the critical machines first using forward, backward and ‘forward & backward’ scheduling techniques, which is a different approach from other scheduling systems commercially available. The “scheduling around bottleneck” approach aims to utilize the bottleneck machines to the maximum and leads to higher output. Master of Science

This report presents a study on the PMC's current supply chain and proposals to improve it. In this report, factors affecting the supply chain are also discussed. Master of Science (Logistics)

This dissertation aims to show how a multi national company reengineered its supply chain to improve its customer service and reduce the supply chain costs to deliver value to the shareholders and customers. This dissertation will study this supply chain reengineering strategy and illustrate the details of the strategy implementation. Master of Science (Logistics)

The report evaluates the impact of the e-Procurement in Apple Computer’s value chain. The focus is on the forecasting and transactions for direct procurement and to provide a common platform between Apple and the suppliers irrespective of the internal systems used by small, medium or large suppliers. Master of Business Administration (Management of Technology)

The mathematical models of Pre VMI and VMI scenarios are developed and analysed. These models are developed to quantify the total cost of maintaining the inventory. Besides, the models are enhanced by considering the probabilistic demand, the service as well as the bullwhip effect. Master of Engineering (MPE)

The objective of this research is to provide valuable insights as well as assist in the selection of an appropriate information exchange mechanism for a particular supply chain. In addition, this research will evaluate hybrid information exchange mechanisms which are possible solutions to achieve higher level of service and simultaneously reduce the level of inventory held by the supply chain. Master of Engineering (MPE)

The research objective is to develop strategies to reduce the complexity needed top create a production-planning simulation model of a semiconductor facility. The thesis is that production planners at high level do not see all the details to develop an effective plan Simplifying the input data of a model as well as decreasing the number of elements in a model can reduce complexities of a model. Master of Engineering (MPE)

The objective of this project is to design and develop a prototype throughput management tool for semiconductor manufacturing. Master of Science (Computer Integrated Manufacturing)

This reports provides an overview of the evolution and purpose of Vendor Managed Inventory (VMI), and the various models available and implemented in the industry today. The report aims to highlight the challenges of suppliers supporting the PC industry in particular. Master of Science (Logistics)

A national survey to evaluate the use of Information Technology in the local logistics industry (Depot Operations) was conducted in April 1999 by the National Computer Board (now known as Info-Communication Development Authority of Singapore) together with the Trade Development Board and the Nanyang Technological University. The common perception is that this industry is generally slow in adopting new Information Technology to improve performance of the logistics operations. It is also perceived that the Government-linked Companies (GLCs) are more receptive to IT implementation than the small local enterprises. Master of Science (Logistics)

This project aimed to provide real time information required for management decisions and planning. It has to start with integration of various processes and controls over manufacturing shop-floor (semi-conductor sector), to achieve common the critical success factor of the organization. Master of Science