当前位置:首页 >> 交通运输 >>

Cost and Benefit Analysis of RFID

Notes and Comments

Cost-Benefit Analysis of RFID Application in Apparel Retailing for SME: A Case from Taiwan
To date, nine out of ten major global retailers, including Walmart, Carrefour, Home Depot, Metro, TESCO, Kroger, and Costco, have planned to use RFID technology to build more convenient retailing environments and to improve warehouse logistical efficiency in terms of reduced out-of-stock rates, enhanced replenishment efficiency, and better inventory management. Globally renowned apparel retailing leaders such as Prada, Gap, Benetton, Levis, and Gardeur, also greatly value the benefits of applying RFID technology (Roberti 2006). The product characteristics of the apparel retail industry constitute a significant seasonal demand, a strong requirement for individualized services, rapid updates, a tremendous time-tomarket pressure, and uncertainty in customer preferences and estimated sales forecasts. These characteristics result in inventory inaccuracy (Bruce et al. 2004). By specifically addressing the characteristics of apparel retailing, the globally renowned leaders are depending on the application of RFID technology to use the advantage of real-time information for making better decisions, as well as boosting the performance of supply chain management from warehouse to sales. Goods will have a better ‘‘information transparency,’’ effectively reducing the operational costs of store management (Moon and Ngai 2008).
Mr. Wen is Ph.D. candidate, Department of Industrial Engineering and Management, National Chiao Tung University, Hsinchu, 30050, Taiwan; e-mail dj.wen@yahoo.com.tw. Ms. Chang is assistant professor, Department of Industrial Engineering and Management, National Chiao Tung University. Mr. Chang is assistant professor, Department of Management Sciences, R.O.C. Military Academy, Kaohsiung 830, Taiwan.

The apparel retail industry does not currently show a great deal of willingness to introduce RFID technology. This may be due to factors such as the heavy investment required for the required hardware and software, a less than 100 percent reading rate, and unresolved privacy issues (Hill 2004). When these issues are coupled with an inferior budgetary allowance for new operational technologies when compared with foreign global leaders, a lack of ability to integrate the technology into supply chains, and a lack of power to bargain for better prices, the cases of RFID application in Taiwanese apparel retailing are very limited. Among the few experimental projects sponsored by Taiwan’s government, the success rate of converting RFID use into actual operations is very low. How can RFID technology be adopted for large-scale applications? Most research on RFID applications in apparel retailing has focused on the study of global leaders, using indepth interviews or questionnaire survey approaches (Moon and Ngai 2008). In order to break the myth that only globally renowned leaders can enjoy the benefits of RFID technologies, this study investigates a case drawn from a small-to-medium-sized apparel retailer that has implemented RFID applications in Taiwan. We use a different approach by using analytical simulation methods to study the differences between operations before and after introducing RFID technology and by examining whether the introduction of RFID produces expected benefits. The results should serve as a reference for the future introduction of RFID technology. The rest of this article is organized as follows. Section two briefly introduces RFID




technology and reviews its application in apparel retailing. Section three describes the background of the company selected for this study and carries out a simulation of operational processes after the introduction of RFID. Section four then analyzes the costs and benefits of RFID technology, as well as the impact on the business model after introducing RFID technology. Finally, the fifth section discusses the findings of this study and makes conclusions and recommendations. RFID TECHNOLOGY AND ITS APPLICATIONS IN APPAREL RETAILING In relation to RFID applications in apparel retailing, Marks and Spencer (M&S), a wellknown British apparel retailer, has applied RFID technology to clothing products for men and women in its clothing stores, hoping to provide customers with better-quality RFIDenabled automated services to avoid situations when customers cannot find the right goods. At present, M&S clothing stores use RFID tags to track and control goods dynamically, from upstream clothing factories through the middle-stage distribution centers (DCs), and downstream to the destination stores. According to M&S, the inventory time of clothing for men and women has shortened from eight hours to one hour after using RFID tags, an 88 percent increase in efficiency. In an experiment with three suppliers that involved 400,000 items, the benefits achieved included enhanced inventory accuracy, increased store sales turnover, greater interaction between staff and customers, and a better grasp of customer orders (James 2007). GAP manages its best-selling jeans through RFID identification, which allows real-time tracking of different styles in terms of sales and purchase status. By clearly grasping realtime information of goods in the supply chain, GAP is able to have a quicker understanding of market trends to make better procurement decisions (Marshall 2006). According to its internal statistics, GAP has gained a 20 percent increase in its sales turnover since applying RFID technology to apparel tracking and management. According to Falabella, a large-scale distributor in Latin America, their first round of RFID application programs has achieved remarkable

results (O’Connor 2007). Through RFID, Falabella enjoys a 98.4 percent accuracy rate in its inventory, while store staffs achieve a higher operational efficiency, with each employee being able to take an inventory of up to 9,000 items per hour. The company also noticed a 25 percent reduction in its out-of-stock rate for RFID-affixed clothing items. Falabella has initiated a plan to expand its second-round RFID application program to include more items. CASE ANALYSIS Introduction to the Case Company This study investigates a real case drawn from an apparel retailer in Taiwan (hereafter to be referred to as Company A). Company A is a small-to-medium-sized enterprise (SME). Its main business is acting as an agent for over ten globally renowned brands, with products mainly focused on leisure apparel and accessories for both men and women. The 2007 sales revenue for the company’s commissioned apparel brands reaches NT$700 million, and the annual growth rate is expected to be over 15 percent for the next three years. This study uses simulation and cost-benefit approaches to explore whether it is cost-effective for Company A to invest in RFID technology. Company A carries medium- and highpriced apparel in its main product lines, with the average unit price over New Taiwan Dollar (NT) $2,000. The product line features limited quantity and rich varieties that, together with a high seasonality of apparel products, positively correlate company revenue with inventory management efficiency. In order to achieve high efficiency in inventory management, Company A has made a certain amount of investments in its information technology (IT) systems, which include point-of-sale (POS) systems, enterprise resource planning (ERP) systems, and platforms for business-to-business (B2B) services. With the increased number of agency brands in recent years, there is a dramatic increase in the frequency of receiving and shipping of goods in Company A’s DC, reaching an annual total of over 450,000 pieces. Also, with more than 70 exclusively owned stores in Taiwan, Company A has to constantly improve its efficiency in goods receiving and shipping, and to




Table 1. Basic Information for Company A Distribution center Number of stores in Taiwan Type of goods Goods characteristics One RFID application environment RFID tag application DC and stores

More than 70

Item level

Average unit price of over NT$2,000 goods ? Limited quantity and rich varieties in products, with high seasonality Men’s & women’s apparel ? Being agents for over ten foreign brands creates a high frequency of goods shipped into and out of the DC (over 900,000 pieces annually); varieties of items result in poor inventory management. ? Highly concentrated arrival time for new items coupled with limitation of manpower and storage space, the company is severely impacted by each new item’s time-to-market as each container of new items takes a long time in the DC, from receiving the goods to shipping them. brand, arriving in a highly concentrated period of time, have to be processed with limited manpower in a limited space. In order to reduce the time required for receiving and shipping goods, Company A provided RFID tags for the manufacturer to affix to each item of clothing before shipment. Company A set up an RFID system (including the fixed gate reader and handheld reader) in the DC to replace the existing barcode system, in order to boost the efficiency of receiving and shipping by the DC by at least 50 percent. The conceptual diagram of Company A’s RFID operational process is shown in Figure 1. To assess the efficiency of goods received and shipped in the DC after RFID technology is introduced by Company A, this study constructed simulation models for Company A’s existing receiving and shipping processes in the DC and the processes after adopting RFID technology, in order to analyze the changes in time spent and input-output benefits generated. The simulation models were built by using eM-Plant software. Using this software, the user can analyze various indicators in a production system, including productivity, capacity utilization, worker load, and the smoothness of the logistics process. This article will describe in detail the variations in the receiving and shipping of goods, and cost-benefit analyses

Current bottleneck

increase its speed in shelving new products. However, Company A is limited by manpower and storage space, which severely constrains each new item’s time-to-market, especially when new items arrive from overseas in highly concentrated time periods. Each container of new items takes a long time between receiving the goods in the DC to shipping them. Company A has started to evaluate an efficiency increase by introducing RFID technology into its DC to streamline goods-receiving and shipping operations. Company A’s basic information regarding this RFID application is summarized in Table 1. Modes of Operation Being a brand agent with all its apparel manufactured by plants abroad, Company A has to provide all the existing barcode tags in advance so they can be affixed by the manufacturers onto each item. Upon arrival of goods, the DC staff has to go through the goods using a barcode reader to read each item. For a typical arrival of a maximum of 30,000 new items (equivalent to approximately 750 cartons), which are handled by one person, it will take about twelve days to finish the process from the moment the goods enter the DC. Indeed, such a task becomes an ordeal for a time-pressured apparel retailer when new items of each




Figure 1. Conceptual Diagram for Company A’s RFID Operation

before and after the introduction of RFID technology into the DC. Process and Operation Time after Introducing RFID Technology into the DC Being a brand agent, Company A’s DC reaches its busiest state after the arrival of new items. The monthly average volume received and shipped by the DC in the busy season reaches over 90,000 pieces. Taking into account such factors as whether the logistics expenditure is economical, Company A has to arrange for the goods-receiving volume to average between 20,000 and 30,000 pieces. Operating under the existing barcode system, Company A’s DC has to, upon receiving each batch of new goods, manually verify item numbers and quantities of goods against the delivery document. It takes, on average, 5.68 days for one operator to verify 30,000 pieces of goods received and 6.62 days for one operator to finish shipping these goods. After adopting RFID technology for the receiving and shipping processes to the DC, the work flows are required to update accordingly. Figure 2 and

Figure 3 illustrate the work flows for the receiving and shipping processes, respectively, before (As-Is) and after (To-Be) adopting RFID technology. The working hours required for processing a container (with 30,000 pieces or 750 cartons on average) were obtained and are shown in Table 2, assuming the number of operators responsible for the receiving and shipping processes in the DC is one and two, respectively (the DC altogether has six employees). It can be seen from the simulation results that, after introducing RFID technology to the DC, the time required for one operator to process the receiving and shipping of 750 cartons of apparel on a carton basis is 5.58 days, a 55 percent saving from the 12.31 days required before adopting RFID technology. With two operators working together, the time required for processing 750 cartons of apparel shortens from 6.83 days to 2.9 days, a 57 percent time saving. From the simulation results, it is obvious that the time saved by adopting RFID technology into the DC meets the key performance indicator (KPI) of at least a 50 percent performance




Figure 2. Receiving Process for Company A’s DC

enhancement for the DCs receiving and shipping operations with one or two operators. In addition to considering the goods received and shipped by the DC of Company A, the RFID system has also brought significant benefits for the inventory counting process. The inventory that is stored in Company A’s DC averages 40,000 pieces, and the inventory

counting process will implement its first and second counts twice a year, and a sampling count once every month. When implementing the first count, the products that will be included in the inventory counting process first need to be found and sorted. Operators will then compare the card of the storage site and, after confirming the correct site, write down




Figure 3. Shipping Process for Company A’s DC




Table 2. Benefit Analysis of Company A’s DC in Terms of Time Saved after Introducing RFID Technology Number of Operators Operation Process Receiving 1 Shipping Receiving and shipping Receiving 2 Shipping Receiving and shipping Before Introduction of RFID 5.68 days 6.62 days 12.31 days 2.84 days 3.99 days 6.83 days After* Introduction of RFID 1.53 days 4.04 days 5.58 days 0.49 day 2.41 day 2.90 days Benefit after Introduction of RFID 73.03 percent time saved 38.94 percent time saved 54.69 percent time saved 73.03 percent time saved 38.94 percent time saved 57.48 percent time saved

*: Receiving and shipping of goods are in the quantity of 750 cartons (or 30,000 piece), with a tag reading accuracy of 95 percent.

the product serial numbers and quantity on the inventory form. In addition, such operations will take one week (five working days) to complete the entire process. The second count will be implemented immediately after the completion of the first count, and will only take one working day (eight hours) to complete the entire process; as for the sampling count, it is conducted once every month and needs one working day (eight hours) to complete the entire process. Moreover, a simulation analysis will be conducted in accordance with the processing time of different operating procedures, and the result is significant in that after the introduction of RFID technology, the barcodes on each item of clothing no longer need to be read one by one; thus, within a two-person operation, the first count time of 40,000 items clothing is cut from five working days to 2.19 working days, with the time-saving ratio of 56.2 percent; the second count time is reduced from one working day to 0.25 working days, with the time-saving ratio of 75 percent; and the time of the sampling count is decreased from one working day to 0.259 working days, with the time-saving ratio of 75 percent. Therefore, within a two-person operation, the time of conducting inventory counting is reduced from seven working days to 2.69 working days, with the time-saving

ratio of 62 percent. The results are compiled in Table 3. BENEFIT ANALYSIS OF INTRODUCING RFID TECHNOLOGY Based on the obtained simulation results described above, introducing RFID technology into Company A’s DC created significant benefits in terms of time saved for goods receiving, shipping, and carrying out inventories. The amount of time saved depends on the process as well as the reading rates and number of operators. In this study, an optimistic estimation of a 95 percent reading rate with two operators was used for the benefit analysis to calculate the total benefits that Company A could gain after introducing RFID technology into its DC. Other than the time saved in receiving and shipping operations, additional benefits included the fact that new items could be shelved earlier in stores for customers, the product could be launched ahead of time, and the idle time during the inventory period could be shortened. The benefits for introducing RFID technology into Company A’s DC are described in terms of costs and benefits as follows. Cost Analysis for RFID Introduction In order to reduce the time required for receiving and shipping goods, Company A provided RFID tags for its manufacturers to affix




Table 3. Benefit Analysis in Time Saved for Inventory Counting after Introduction of RFID System into Company A’s DC Number of Operators Inventory Operation First count 2 Second count Sampling count Total Before Introduction of RFID 5 days 1 days 1 days 7 days After Introduction of RFID 2.19 days 0.25 day 0.25 days 2.69 days Benefit after Introduction of RFID 56.2 percent time saved 75 percent time saved 75 percent time saved 61.57 percent time saved

Table 4. Costs for Introducing RFID Technology into Company A’s DC Unit: NT$ Item Hardware cost Software cost Cost Variable cost Fixed cost Fixed cost Item RFID tags Printer Handheld reader Gate reader Middleware System integration Total cost Unit Price Qty Total Subtotal $10 450,000 $4,500,000 $4,500,000 $150,000 1.00 $150,000 $100,000 2.00 $200,000 $850,000 $250,000 2.00 $500,000 $1,000,000 1.00 $1,000,000 $2,000,000 $1,000,000 1.00 $1,000,000 $7,350,000 Benefit Analysis after Introducing RFID Technology The labor cost saved after introducing RFID technology into Company A’s DC is summarized in Table 5. According to data shown in Table 2, an average of 3.93 (=2.84+3.99-0.492.41) working days was saved for receiving or shipping goods after introducing RFID technology, assuming 750 cartons (30,000 pieces) can be processed by two operators at a 95 percent reading rate. From Table 5, with Company A’s data showing an average monthly volume of receiving and shipping during the busy season to be 45,000 items each, the average monthly saving in working days is 5.895 days (3.93 days * 45,000/30,000); with average low-season monthly receiving and shipping volumes being 15,000 items each, the average monthly savings in working days is 1.965 days (3.93 days * 15,000/30,000). In addition, with regard to inventory counting, the total time saved from the first and

to each item of clothing before shipment. Company A set up an RFID system (including the fixed gate reader and handheld reader) in the DC to replace the existing barcode system. The estimated costs incurred in this application for Company A are shown in Table 4. In terms of RFID tags, no recovery cost is considered, assuming the company did not intend to recover them for re-use. With each piece of clothing affixed with a UHF RFID tag, which has an estimated cost of NT$10 per piece, and with the current total number of goods received annually in the DC being 450,000 pieces, the estimated total variable cost for RFID tags is NT$4,500,000. Other hardware required includes a printer, handheld readers, and gate readers. The software includes middleware, as well as system integration costs. It is estimated that a total cost of NT$7,350,000 is required, of which NT$4,500,000 is a variable cost and NT$2,850,000 is a fixed cost.




Table 5. Labor Cost Saved after Introducing RFID Technology into Company A’s DC Unit: NT$ Item Busy season Low season Inventory counting Number of Months 9 months 3 months Avg Qty of Goods in/out per Month 45,000 pieces 15,000 pieces Avg Days Saved per Month 5.895 1.965 Avg Working Weeks Saved 1.179 0.393 3.224 Total working months saved Average monthly salary at the DC Total annual operation expenses saved Avg Working Months Saved 0.29475 0.09825 0.806 Total Working Months Saved per Month 2.65275 0.29475 0.806 3.7535 $30,000 $112,605

Note 1: Average time saved for both receiving and shipping processes is 3.93 working days when 30,000 pieces are processed by each operation. Note 2: Number of working days per week is five days. Note 3: Number of working weeks per month is four weeks.

second counts (each to be carried out twice a year) would be (5-2.19)*2+ (1-0.25)*2=7.12, and (1-0.25)*12=9 would be the time saved in the sampling count (once every month); thus, the average number of working days that are saved every year would be 16.12. An average monthly salary of NT$30,000 should be assumed for operators working in Company A’s DC. Each operator is required to work eight hours a day and five days per week, or twenty days per month. An optimistic estimate that is based on the annual receiving and shipping of 450,000 pieces each would generate a saving of NT$112,605 in labor costs. Extra Benefits and Constraints According to the data shown in Table 2, the average time saved for processing (both receiving and shipping) the assumed 750 cartons of goods with a 95 percent reading rate carried out by two operators is estimated to be 3.93 days, which translates into 0.786 weeks (3.93 working days/5 working days per week) of earlier shelving. According to Little’s law (Little 1961), the key benefits for a reduced cycle time is a shortened time-to-market and speeded-up flow of goods. A reduced cycle time also ensures a slow price drop for finished

products before they hit the market, so maintaining the average selling price at a higher level. According to apparel retailers, after introducing RFID technology in accordance with the characteristics of apparel retailers, the early shelving of new items can advance the productselling period. Such an effect will be revealed in the last stages of the extended selling time. To customers, the early effects of apparel being shelved earlier do not make much difference along the time axis; but the extended life cycle due to introduction of RFID technology will bring in lucrative effects later on, assuming that each sales session comprises an average of three months (corresponding to each season) and that the profit of early shelving will occur during the last stage (week) of each session. Moreover, in the case of Company A, the semi-annual inventory counting process requires the DC to reserve manpower one month in advance to start counting items of low shipping frequencies. The counted items are boxed and affixed with labels indicating the contents and associated quantities. These items cannot be shipped unless an urgent request from stores arrives or the semi-annual inventory counting process is completed. The situation is further




complicated by the variety of items under the control of the DC. Limited manpower and storage space for both receiving and shipping processes, in parallel with the inventory counting process, creates a situation where the physical inventory differs from inventory records. In general, as inventory processing takes up manpower, introducing RFID technology can enhance inventory counting efficiency as well as accuracy. Finally, one should note that the reading accuracy of handheld readers is affected by the moving speed of the operator who controls the reader and by interference from metallic shelves. Therefore, RFID has yet to overcome some technical issues of low reading rates in the case of overlapped RFID tags (which have only a 90 percent reading rate) and for interference caused by metals such as apparel accessories and aqueous substances like perfumes. CONCLUSIONS The purpose of this research was to find out how RFID technology can be introduced into applications that currently lack RFID, including large-scale applications. Furthermore, as applications of RFID technology in apparel retailing do not follow a standard model, a broad customization of system services is required for different operational environments. In order to break the myth that only globally renowned leaders can enjoy the benefits of RFID technologies, this study picked Company A, an apparel retailer in Taiwan, for a case analysis. By using the eM-Plant system simulation software to model the operations after RFID technology was introduced into Company A’s DC, we found that over 54 percent in operation time was saved in one- or two-person operations, which meets the KPI of at least 50 percent in performance enhancement for the DC’s goods receiving and shipping operations, as expected at the evaluation stage of RFID technology introduction. In terms of input costs, the hardware and software necessary for introducing RFID technology into Company A’s DC includes RFID tags, tag printers, handheld readers, gate readers, middleware, and system integration, totaling NT$7,350,000 at current market prices. In terms of benefits gained, other than operation

time saving, which translates into lower labor costs, savings in goods receiving and shipping time will speed up the efficiency of taking a physical inventory and shorten the period of goods’ idle time before shipping, so improving Company A’s highly concentrated arrival time for new products. It is estimated that the timeto-market or early shelving of new items is about one week earlier after the introduction of RFID technology (the actual number being 0.786 weeks) and yields extra benefit from the early shelving of new items. The benefit analysis for Company A shows that for a small-to-medium-sized apparel retailer adopting RFID technology, the direct benefits are negative, when simply weighing the fixed and variable costs of RFID implementation vs. the benefits from manpower savings. However, if additional benefits are also considered, such as the benefits from enhanced inventory counting efficiency and accuracy, the decrease in time-to-market for new products, and the benefits from being able to shelve goods sooner after their acquisition, then RFID applications may be profitable. REFERENCES
Bruce, M., L. Daly, and N. Towers (2004), ‘‘Lean or Agile: A Solution for Supply Chain Management in the Textiles and Clothing Industry?’’ International Journal of Operations and Production Management, 24(2), pp. 151–170. GS1 Germany and WP7 partners (2007), ‘‘Supply Chain Management in the European Textile Industry,’’ BRIDGE, <http://www.bridge-project.eu/data/File/BRIDGE percent20WP07 percent20Textile percent20Industry percent20- percent20Requirements percent20analysis.pdf>. Hill, S. (2004), ‘‘Apparel Logistics: New Technology, Trends, and Industry Developments – Management Briefing: RFID,’’ Just–Style, July, pp. 12-15. James, S. (2007), ‘‘How M&S Is Using RFID to Improve Customer Service and Retail Productivity’’ RFID Journal, <http://www.rfidjournalevents.com/files/main/ Tues_1645_Stafford.pdf>. Moon, K.L. and E.W.T. Ngai (2008), ‘‘The Adoption of RFID in Fashion Retailing: A Business Value-Added Framework,’’ Industrial Management and Data Systems, 108(5), pp. 596–612. Little, J.D.C. (1961), ‘‘A Proof for the Queuing Formula,’’ Operations Research, 9, pp. 383–387. Marshall, K. (2006), ‘‘Missing: Specialty Apparel Retailers,’’ RFID Journal, <http://www.rfidjournal.com/ article/articleview/2531/1/82/>. O’Connor, M.C. (2007), ‘‘Falabella Plans Second ItemLevel RFID Pilot,’’ RFID Journal, <http:// www.rfidjournal.com/article/articleview/3585/>. Roberti, M. (2006), ‘‘RFID is Fit to Track Clothes,’’ Chain Store Age, 82(5), p.158.

Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.