Optimizing Decision Making in the Apparel Supply Chain Using Artificial Intelligence (AI)

From Production to Retail

Calvin Wong, Z. X. Guo, S Y S Leung

• Publisher: Woodhead Publishing
• Year: 2013
• Print ISBN: 9780857097798
• eBook ISBN: 9780857097842
• Topic: Business
• Subtopic: Fashion & Textile Industry
• Index: Business

Practitioners in apparel manufacturing and retailing enterprises in the fashion industry, ranging from senior to front line management, constantly face complex and critical decisions. There has been growing interest in the use of artificial intelligence (AI) techniques to enhance this process, and a number of AI techniques have already been successfully applied to apparel production and retailing. Optimizing decision making in the apparel supply chain using artificial intelligence (AI): From production to retail provides detailed coverage of these techniques, outlining how they are used to assist decision makers in tackling key supply chain problems. Key decision points in the apparel supply chain and the fundamentals of artificial intelligence techniques are the focus of the opening chapters, before the book proceeds to discuss the use of neural networks, genetic algorithms, fuzzy set theory and extreme learning machines for intelligent sales forecasting and intelligent product cross-selling systems.- Helps the reader gain an understanding of the key decision points in the apparel supply chain- Discusses the fundamentals of artificial intelligence techniques for apparel management techniques- Considers the use of neural networks in selecting the location of apparel manufacturing plants

1

Understanding key decision points in the apparel supply chain

W.K. Wong, The Hong Kong Polytechnic University, China

Abstract:

In the apparel supply chain, a range of key decisions are always faced by apparel enterprises. These decisions, including site selection for establishing manufacturing plant, production planning and scheduling, line balancing, sales forecasting, etc., rely on the experience and subjective assessment of management and decision makers. As the apparel industry is characterized by short product life cycles, volatile customer demands and tremendous product varieties, such decisions have become more complex. This chapter will discuss these key decision points.

Key words

plant location

production planning and scheduling

marker planning

cut order planning

spreading

cutting

line balancing

sales forecasting

cross-selling and up-selling

1.1 Introduction

Apparel manufacturers and retailers in the fashion industry face a range of key decisions, including selection of plant locations, production planning and scheduling, marker planning, cut order planning, apparel assembly line balancing, retail sales forecasting and marketing. Traditionally, such decisions depended on the experience and judgement of key staff. However, as the market has shifted to short production runs to meet rapidly changing customer demands, and costs have been squeezed in favour of just-in-time production methods, these decisions have become more complex. At the same time, production has become more automated and integrated, allowing greater control of the supply chain. In this chapter, key decisions in the apparel supply chain will be discussed.

1.2 Selection of plant locations

Apparel manufacturers’ direct investment and joint ventures in developing regions have grown rapidly in the past few decades. The choice of plant locations for foreign direct investment is an important decision. Non-optimized selection can adversely affect a plant’s performance in terms of productivity, manufacturing and logistics costs. Selection of a proper plant location is thus crucial. In the case of establishing overseas plants, apparel manufacturers should consider costs, profits and other intractable factors, such as social environment, political stability, legality, technology, and micro-environmental factors, including customers, competitors and suppliers. Most manufacturers have difficulties in decision making due to vague and subjective measures, particularly for variables not represented by objective values, such as country risk and community facilities. The decision of plant locations thus mostly relies on the intuition and assessment of manufacturers.

1.3 Production scheduling and assembly line balancing control

The current competitive market environment causes difficulties in scheduling and line balancing control in the modern apparel industry.

1.3.1 Production scheduling

In today’s apparel industry, fashion products require a significant amount of customization due to differences in body measurements, diverse style preferences and replacement cycles. Apparel manufacturers are usually given a short production lead-time, tight delivery dates and small quantities with frequent style changes. To cope with the increasing demand for product customization, the quantity of garments per production order tends to be smaller and thus the number of production orders is higher.

It is necessary for apparel supply chains to be responsive to the ever-changing fashion markets by producing smaller jobs in order to provide customers with timely and customized products. Because of ever-increasing global market competition, apparel manufacturers have to improve their production performance continuously to be more competitive. Effective production planning and scheduling (PPS) plays a significant role in maximizing resource utilization and shortening the production lead time. As PPS decisions mostly rely on production planners’ ad hoc assessment and intuition, they may not be consistent or optimized even under similar conditions. All this makes it more difficult for manufacturers to make effective PPS decisions.

In the real-life production environment, various uncertainties often occur, such as customer orders and processing time. An estimate not conforming to industrial practice can lead to an unsatisfactory scheduling solution. Without considering uncertainties, it is difficult to produce an optimized production schedule and thus hard to achieve optimal performance. For example, if a schedule fails to factor in possible future orders, rush orders can disrupt the production of orders which have already been scheduled.

Some commercial PPS systems only provide a platform for conducting PPS arrangements, but cannot automatically provide scientific and optimized solutions. PPS decisions in the apparel industry still rely heavily on production schedulers’ experience, intuition and assessment rather than a scientific and systematic approach.

1.3.2 Assembly line balancing control

The assembly sewing process is the most labour-intensive part of apparel manufacturing. The progressive bundle unit system is common in sewing room design. Recently, many manufacturers have installed unit production systems as a means to improve efficiency and effectiveness. Assembly involves a set of workstations in which a specific task of a pre-defined sequence is processed.

In order to achieve a balanced line before production, sewing line supervisors usually assign one or more sewing operatives to each task based on the standard time required to complete the task. However, it is difficult to achieve a perfectly balanced line because the production rate of each workstation is different. Imbalance occurs due to various factors, including fluctuation in operative efficiency, frequent changes of product styles, order size, prior experience and unexpected factors, such as absenteeism and machine breakdown. Line balancing control is required to smooth away bottlenecks.

The balance control of an apparel assembly line relies heavily on the shop-floor expert’s knowledge, experience and intuition. The effectiveness of a decision depends on the subjective and ad hoc assessment of production line supervisors. Small order size and frequent changes of styles can make the matter even worse for optimal production control. With the recent development and adoption of real-time shop-floor data capture systems, real-time production statistics and progress reports can be generated to assist production line supervisors in line balancing control and bottleneck elimination. However, their decisions may not be consistent even under similar conditions and may thus be non-optimal.

1.4 Cutting room

The key decision points in the cutting room of apparel manufacture include cut order planning, marker planning, and spreading and cutting scheduling.

1.4.1 Cut order planning

In apparel supply chains, fabric is the single largest contributor to garment costs. Approximately 50–60% of manufacturing costs can be attributed to fabric. Apart from fabric, labour and factory operation costs have also been continuously increasing while selling prices of apparel products have been falling, which presents a great challenge to apparel manufacturers to adopt quick response strategies to manufacture and deliver apparel products to retailers while maximizing fabric utilization rates (i.e. minimizing material costs) and minimizing labour and manufacturing costs.

Cut order planning (COP) is the first stage in the production workflow of a typical apparel manufacturing company upon receiving a production order from a client (Fig. 1.1). It is the process to determine the number of markers needed, the number of garment sizes in each marker, and the number of fabric plies to be cut from each marker. Markers are the output of marker planning, which is the operation following COP. Figure 1.2 shows a marker planning process using commercial computing to arrange all patterns of component parts of one or more garments on a piece of marker paper (Fig. 1.3). The third operation is fabric-spreading, in which fabric pieces are superimposed to become a fabric lay on a cutting table (Fig. 1.4). The last operation is fabric-cutting. Garment pieces are cut out of fabric lays according to the pattern lines of component parts of one or more garments on the marker, and then assembled by the sewing department as a finished garment.

1.1 Schematic workflow of activities of a fabric-cutting department of a typical apparel manufacturing company.

1.2 Example of marker paper.

1.3 Marker planning process using commercial computing software.

1.4 Fabric lay composed of fabric plies after spreading.

COP, the most upstream activity in apparel manufacturing, plays a significant role in affecting fabric and manufacturing costs in the cutting department. Based on the requirements of customer orders in terms of style, quantity, size and colour, it seeks to minimize total production costs by developing cutting orders with ...

Table of contents

• Cover image
• Title page
• Table of Contents
• Copyright
• Woodhead Publishing Series in Textiles
• Preface
• Acknowledgements
• Chapter 1: Understanding key decision points in the apparel supply chain
• Chapter 2: Fundamentals of artificial intelligence techniques for apparel management applications
• Chapter 3: Selecting the location of apparel manufacturing plants using neural networks
• Chapter 4: Optimizing apparel production order planning scheduling using genetic algorithms
• Chapter 5: Optimizing cut order planning in apparel production using evolutionary strategies
• Chapter 6: Optimizing marker planning in apparel production using evolutionary strategies and neural networks
• Chapter 7: Optimizing fabric spreading and cutting schedules in apparel production using genetic algorithms and fuzzy set theory
• Chapter 8: Optimizing apparel production systems using genetic algorithms
• Chapter 9: Intelligent sales forecasting for fashion retailing using harmony search algorithms and extreme learning machines
• Chapter 10: Intelligent product cross-selling system in fashion retailing using radio frequency identification (RFID) technology, fuzzy logic and rule-based expert system
• Index