RFID IN THE SUPPLY CHAIN: HOW TO OBTAIN A POSITIVE ROI

The Case of Gerry Weber

Christoph Goebel, Christoph Tribowski, Oliver G

unther

Institute of Information Systems, Humboldt-Universit

at zu Berlin, Spandauer Str. 1, 10178 Berlin, Germany

Ralph Tr

oger, Roland Nickerl

Gerry Weber International AG, Neulehenstr. 8, 33790 Halle/Westfalen, Germany

Keywords:

RFID, Supply chain management, Case study, Apparel industry.

Abstract:

Although the use of Radio Frequency Identiﬁcation (RFID) in supply chains still lags behind expectations, its

appeal to practitioners and researchers alike is unbowed. Apart from technical challenges such as low read

rates and efﬁcient backend integration, a major reason for its slow adoption is the high transponder price.

We deliver a case study that investigates the ﬁnancial, technical and organizational challenges faced by an

apparel company that is currently introducing item-level RFID to monitor their supply chain. The company

has developed an implementation strategy based on cross-company closed-loop multi-functional use of RFID

transponders. This strategy leads to a positive ROI in their case and could serve as an example for other

companies considering the introduction of item-level RFID.

1 INTRODUCTION

Radio Frequency Identiﬁcation (RFID) is a technol-

ogy, whose impact on supply chain management has

been rising steadily. Apparel companies have shown

a particularly high interest in the RFID-based track-

ing of their products to improve logistical operations

(Loebbecke and Huyskens, 2007). This high inter-

est can be explained by the fact that apparel has a

number of properties that make the application of

RFID tags attractive. First, reading out data on RFID

transponders applied to cartons containing apparel or

the labels attached to garments is already feasible

with today’s RFID reader technology. Second, most

clothing items have higher margins as compared to

fast moving consumer goods. On the one hand, this

makes tag prices appear a less crucial ﬁnancial con-

straint. On the other hand, high product margins in-

crease the attractiveness of RFID-based theft preven-

tion. Third, being able to monitor the movements of

items on the shop ﬂoor has long excited the imagi-

nation of marketing professionals in the apparel sec-

tor. For instance, innovative retail applications such

as the RFID-enabled changing room are under consid-

eration among fashion retailers. However, the com-

mercial success of any application directly interfac-

ing with the end-consumers in a proactive way cru-

cially depends on their acceptance. In the case of

RFID-based applications, upfront technology accep-

tance has turned out to be rather low due to fears re-

garding privacy intrusion (G

unther and Spiekermann,

2005).

The initial euphoria about RFID’s potential has re-

cently made way for a more down-to-earth view of its

beneﬁts in the supply chain. Companies who con-

sider using RFID usually conduct a conservative pre-

liminary analysis of the ﬁnancial impact of such an

investment. These analyses typically focus on three

types of beneﬁt expected from RFID: The reduction

of labour, capital and non-conformity costs such as

costs caused by wrong deliveries. Labour and capi-

tal costs can be decreased by RFID-enabled process

speedups, while non-conformity costs can be reduced

by detecting mistakes made during the distribution

process and taking appropriate action to prevent them.

Given the still relatively high cost of passive RFID

transponders (currently about 7 Eurocents), the use of

disposable transponders on the item level represents a

signiﬁcant increase of variable costs in the short run.

Based on the authors’ knowledge gained in several

RFID projects, the one time use of RFID transpon-

ders to support standard distribution processes (e.g.

Goebel C., Tribowski C., Günther O., Tröger R. and Nickerl R. (2009).

RFID IN THE SUPPLY CHAIN: HOW TO OBTAIN A POSITIVE ROI - The Case of Gerry Weber.

In Proceedings of the 11th International Conference on Enterprise Information Systems - Information Systems Analysis and Speciﬁcation, pages 95-102

DOI: 10.5220/0001967700950102

 SciTePress

picking, packing, shipping etc.) is hardly justiﬁable

in ﬁnancial terms.

There are two interdependent reasons for the slug-

gish item level introduction of RFID: high transpon-

der prices and low demand. In 2006, the biggest share

of RFID transponders produced worldwide (556 mil-

lion) went to products such as ’smart’ cards, keys,

passports and tickets (IDTechEx, 2007). Only 153

million of them were sold for the purpose of iden-

tifying goods including drugs, tools, books, apparel

and other consumer products. Roughly 235 million

transponders went to the identiﬁcation of logistical

units such as packages, cases and pallets. The mar-

ket price of RFID transponders is the pivotal param-

eter of most proﬁtability calculations. The efﬁciency

gains achievable in the logistics operations of organi-

zations such as labour cost savings and prevention of

process errors has to outstrip RFID transponder and

infrastructure costs. Otherwise, the vision of perva-

sive RFID tagging is unlikely to become a reality. Ac-

cording to industry experts, the RFID market is sup-

posed to take off as soon as item level tagging in lo-

gistics applications becomes economically viable. On

the other hand, economies of scale in producing RFID

transponders cannot be fully realized since no large-

scale item level RFID implementations exist so far. In

this case study, we present an approach which could

enable ﬁrst movers from the apparel industry to over-

come this deadlock.

The RFID strategy of our case company is based

on the repeated cross-company and combined use

of RFID transponders. The implementation of their

strategy results in a positive ROI even when using

conservative assumptions with respect to expected

costs and beneﬁts. The supply chain participants

share the transponder cost while realizing RFID’s

full potential in applications along the chain. Used

transponders are fed back into the system by taking

them off sold products and putting them onto the new

ones. In addition to that, speciﬁc inter-organizational

beneﬁts such as the detection of problems in upstream

stages of the supply chain can be realized. This appli-

cation is known as Supply Chain Event Management

(SCEM) and has oftentimes been implicated with the

RFID technology. Compared to singular use of RFID

transponders within one company, their repeated use

across ﬁrm boundaries involves additional technical

challenges and a high coordination effort.

The motivation for this work is to show a feasible

way out of the transponder price deadlock. We an-

alyze the potential of item-level RFID in the fashion

supply chain by conducting a case study in the ap-

parel industry and explore the challenges involved in

implementing cross-company closed-loop RFID ap-

plications.

Our paper addresses the following key research

questions:

• What are the advantages and disadvantages of

the cross-company closed-loop use of RFID

transponders compared to single-company open-

loop usage?

• Which challenges are involved in realizing such

RFID applications on the organizational and tech-

nical level?

• How does the case company plan to deal with

these challenges?

The paper is structured as follows. We begin with a

review of the related literature followed by a justiﬁca-

tion of our methodology. Section 3 contains a descrip-

tion of the case. Section 4 summarizes our ﬁndings

and concludes the case study.

2 RELATED LITERATURE AND

METHODOLOGY

The use of RFID in supply chains has received in-

creasing attention among researcher in the ﬁeld of

information systems and operations management in

recent years (Ngai et al., 2008). While the potential

of transponder technology in manufacturing has long

been acknowledged and also realized in many cases,

the use of item-level tagging in the distribution of ﬁn-

ished products is still rare (Schmitt and Michahelles,

2008). Most of the research conducted on RFID in

distribution so far focuses on the ex-ante assessment

of RFID beneﬁts. Within this stream of research, most

attention has been paid to efﬁciency gains enabled

by RFID (Gaukler and Seifert, 2007). While authors

with an industry or consultancy background usually

strived to be more comprehensive in terms of ben-

eﬁt estimation, academic research has concentrated

on speciﬁc types of beneﬁt but has stayed more gen-

eral (Lee and

Ozer, 2007). (Chappell et al., 2003)

for instance, for the case of a major consumer mar-

ket, came to the conclusion that RFID could save up

to 3% of sales. Academic researchers from the op-

erations management ﬁeld have investigated primar-

ily the impact of more accurate information and in-

formation sharing on supply chain management (Lee

and

Ozer, 2007). Researchers from the information

systems and computer science communities have fo-

cused more on the technical aspects of RFID such as

data management (Chawathe et al., 2004) and security

(Juels, 2006).

Although many RFID beneﬁts in distribution op-

erations have been described and estimated in mone-

ICEIS 2009 - International Conference on Enterprise Information Systems

tary terms, hitherto not much research has been pub-

lished based on actual item-level tagging cases. In

particular, real world challenges have not been dis-

cussed in the information systems community. This

lack of practical research can be partly attributed to

the transponder price deadlock discussed above: since

item-level adoption rates are low, practical experi-

ences with the technology are rare. However, the

existence of ’best practices’ based on real world ex-

periences with new technology applications are of-

tentimes crucial determinants for their adoption (Wu

et al., 2006). One of the main inhibitors of RFID

adoption is seen in technology uncertainty, in particu-

lar the risk of not being able to integrate RFID infras-

tructures with existing ERP solutions (see Figure 1).

Therefore, more research on this issue is warranted.

Especially, applied research taking into account real

world challenges should be of great value to practi-

tioners as well as to researchers in the ﬁeld.

RFID diffusion

Decreasing

tag prices

Technology

uncertainty

Negative impact Positive impact

Figure 1: Inﬂuencing factors on the diffusion of RFID,

based on (Strassner, 2005).

In this paper we use a case study approach to come

closer to answering the research questions stated in

Section 1. Case studies are popular both in opera-

tions management and information systems research

(Voss et al., 2002) since they offer insights into the ac-

tivities and experiences of a particular company. As

we mentioned above, research about RFID applica-

tions in SCM hitherto focusses on the ex-ante esti-

mation of beneﬁts. Given today’s low RFID adop-

tion rates, empirical investigations of the beneﬁts and

challenges involved in the application of RFID in

distribution processes would not be very fruitful at

this time. In contrast to that, the analysis of single

RFID adoption cases already makes sense taoday and

should be the preferred research method until adop-

tion has spread further. Case study research can be

used for exploratory, descriptive and explanatory pur-

poses (Schmitt and Michahelles, 2008). In contrast to

quantitative research methods, case studies serve to

analyze single or multiple cases with respect to sev-

eral dimensions of relevance. The case presented in

this paper serves to explore ﬁnancial, technical, and

organizational challenges involved in the design, im-

plementation and deployment of item-level RFID in

the goods distribution process.

3 CASE STUDY

3.1 Company Proﬁle

The Gerry Weber International AG (GW) is a glob-

ally operating apparel company based in Germany.

GW’s primary business consists of the design and

marketing of women’s fashion. Currently about 82%

of sales are generated by the wholesale channel, i.e.

the lion’s share of GW’s products are sold to end cus-

tomers by fashion retailers. The remaining 18% are

either sold by GW’s own brand stores or GW fran-

chisees. GW plans to increase the revenue generated

by its own shops to at least 50%. With about 2,000

employees, GW realized a turnover of over 507 mil-

lion Euros and an EBITDA margin of 12.2% in the

ﬁscal year 2006/2007, which is well above industry

average. GW’s recent success has been attributed to a

fundamental optimization process which among other

things resulted in the outsourcing of logistics opera-

tions.

While the physical operations such as manufactur-

ing, transportation and warehousing have been out-

sourced, supply chain management activities are or-

chestrated from the GW headquarters. GW sources

globally: 63% of the suppliers are located in the Far

East, another 23% in Turkey and the rest in Eastern

Europe. The focal target of GW’s supply chain man-

agement is to assure the reliable delivery of products

to wholesale customers and their own points of sale.

Logistic operations have recently moved to the centre

of attention because it has to keep pace with a recently

introduced marketing strategy: product life cycles in

the retail channel have been reduced to two weeks in

order to improve the customer buying experience and

increase the average number of store visits. GW has

been working on concepts of RFID usage in their sup-

ply chain for several years. In 2007 the decision to

introduce RFID on the item level was made.

3.2 The Supply Chain

3.2.1 The Flow of Goods

GW products are designed in the GW headquarters.

When the major wholesale customers have placed

their orders, GW routes manufacturing orders to con-

tract manufacturers around the world. Besides a small

number of ’never out of stock’-products (e.g. dark-

coloured business suits), which are replenished on

a regular basis, all products are ordered in one big

rush ahead of their respective selling season and are

produced and distributed according to a predeﬁned

schedule. Suppliers can be distinguished into two

RFID IN THE SUPPLY CHAIN: HOW TO OBTAIN A POSITIVE ROI - The Case of Gerry Weber

groups (large and small) according to the size of pro-

duction orders placed with them: Large suppliers ﬁll

whole sea containers or trucks, whereas the shipments

of small suppliers get consolidated and ﬁlled into con-

tainers at consolidation sites near harbours or airports

by the long-haul transportation providers. Depend-

ing on the location of suppliers, merchandise is trans-

ported by truck, sea or air to distribution centres lo-

cated in Germany by major long-haul transportation

providers. Since long-haul transportation services for

apparel are only provided by a small number of major

logistics companies, the probability that GW uses the

same logistics provider repeatedly is relatively high.

In places with big concentrations of contract manu-

facturers of apparel such as the Shanghai area, long-

haul transportation providers operate large consolida-

tion centres where all products destined for overseas

are assembled to shipment lots.

The distribution centres in Germany are operated

by logistics service providers specialized on the ap-

parel industry. These providers also conduct central-

ized quality assurance (QA) and picking of shipments

destined for wholesale and GW’s brand stores. Since

the consumer segment addressed by GW is very sen-

sitive to the quality of products, GW closely monitors

the QA process in order to prevent negative impacts

on revenue. The commissioning process conducted in

the distribution centres is demanding due to two rea-

sons. Firstly, the array of selling points being served

is very heterogeneous in terms of batch sizes, pack-

aging and accounting requirements. Secondly, GW

reserves the right to interfere in the distribution pro-

cess up to the last minute in order to efﬁciently steer

the distribution of goods. In case of inconsistencies of

the actual supply process with schedules, GW’s man-

agement can take immediate action in order to opti-

mize the allocation of the available garments to stores.

The actual transportation from distribution centres to

stores and between stores is done by several short-

haul transportation providers. GW products can be

divided into two categories: Hanging and lying gar-

ments. Hanging garments are transported on hangers

while lying goods are put into cartons.

3.2.2 The Information Flow

In this section we outline which data is received and

sent by the different members of the supply chain dur-

ing the production and distribution of GW products.

We focus on information objects which are directly

associated with physical items. In particular, this in-

cludes production orders and Advance Shipment No-

tices (ASNs).

Production is initiated by sending a production or-

der to a supplier. The production order contains ex-

act information about the type and quantity of all gar-

ments to be produced. Upon receipt of a production

order, the supplier incorporates it into its production

schedule. The supplier is able to access and modify

the order data objects representing the production or-

ders using GW’s SCM system. Among other things,

production progress is recorded in the form of mile-

stones along the garment production process (dyeing,

cutting, sewing, washing) while the expected time to

ﬁnish an order is constantly updated. Finally, when

a supplier prepares a batch of ﬁnished products for

shipment, an Advance Shipment Notice (ASN) is cre-

ated by GW’s SCM software. This ASN provides

an overview of all items contained in the shipments

prepared to leave the supplier’s site. The items listed

by the ASN can be physically distributed over several

containers and/or truck loads and be packaged in car-

tons and hanging packs. The information contained in

the supplier’s ASN only refers to the item type, not the

individual garments. Without item-level RFID there

is no way to tell which carton contains which items

and inside which container or truck load a certain

hanging pack or carton is shipped. The garment lots

received from the small suppliers are consolidated

into shipment lots by the long-haul transportation

providers. When transportation batches have been de-

termined, the long-haul transportation providers send

an ASN via GW’s SCM system. Changes of trans-

portation status are also recorded by the system. Fig-

ure 2 describes the structure of the logistical data pro-

cessed in GW’s supply chain using an UML notation.

The logistics service providers that operate the

DCs receive the ASNs of the long-haul transporta-

tion providers as soon as they are available within the

SCM system. The DC processes have to be adapted to

the type and quantity of incoming deliveries. There-

fore, the ASN information helps the DC operators to

optimize their processes. For instance, if the ASNs

indicate that a larger than expected delivery will ar-

rive, capacity can be added in order to deal with the

situation. The information which GW receives from

the DC operators includes the quantity of garments re-

ceived, the results of the quality control which is con-

ducted shortly after shipments are received, as well as

the results of the picking process. The deliveries re-

ceived by the DC are counted either manually or auto-

matically on a 100 per cent basis, i.e. the data received

by GW allows them to determine whether the suppli-

ers have shipped the agreed quantity. The information

provided on the quality assurance process gives GW

the chance to react based on the number of production

lots which have not passed the tests. For instance, if

a signiﬁcant number of garments are spoilt, the store

orders need to be reviewed and scheduled shipment

ICEIS 2009 - International Conference on Enterprise Information Systems

Production Order

LSP ASN

Supplier ASN

Long-haul TP ASN

Garment

RFID Tag

Pack

Carton

Hanging Garment

Lying Garment

Label

SKU

Packaging Unit

Transportation Unit Truck

Container

1..n

0..1

1..n

0..n

0..1

1..n

0..n

1..n1

0..n

Figure 2: Structure of logistical data.

quantities have to be revised. The information about

the outcome of the picking process serves to create

the ASN of the logistics service provider (LSP ASN)

which is sent to customers. It is also used to proof

to the short-haul transportation provider how many

items have been handed over at the DC in case gar-

ments are lost on the way.

3.2.3 The Potential of RFID

In the case of Gerry Weber, the main ’hard’ beneﬁt

expected from RFID is the reduction of wrong deliv-

eries. The completeness and timeliness of shipments

will be monitored by the use of RFID scanners at sev-

eral steps in the supply chain. Reaction to delays or

mistakes can therefore be carried out more timely and

accurately in the future; for instance, mistakes made

in the picking processes can be identiﬁed promptly

which improves the chances to correct the mistake be-

fore it can cause problems and additional work. An-

other source of quantiﬁable beneﬁt is the reduction of

administrative overhead at GW’s own stores. For in-

stance, sales employees have to count incoming and

outgoing goods either manually or by scanning the

SKU barcode. With RFID, these processes will take a

fraction of the time and enable employees to pay more

attention to the customers.

A number of RFID beneﬁts expected by GW can-

not be quantiﬁed in monetary terms but have never-

theless been taken into account. In particular, GW’s

general management perceives the introduction of

RFID as a strategic investment since their wholesale

customers may also demand tagged products soon.

The decision to introduce item-level RFID was

made based on an economic analysis which consid-

ered the ’hard’ savings mentioned above, i.e. the re-

duction of non-conformity costs at the DCs as well

as the labour cost savings at the stores. The initial

ROI calculation assumed the use of stand-alone RFID

transponders for one-time use. Since this assessment

resulted in a less beneﬁcial ROI, GW’s management

searched for ways to increase beneﬁts and/or reduce

costs.

3.3 Cross-Company Closed-Loop

Integrated Use of RFID

GW came up with three changes to the initial invest-

ment plan:

• Cross-company RFID infrastructure: RFID data

will be used along the whole supply chain in order

to further reduce non-conformity costs.

• Closed-loop application: RFID transponders will

be covered in plastic hard cases so that they can be

used several times in order to reduce transponder

costs.

• Integration of RFID with EAS: beside the RFID

transponder, every hard case will also have Elec-

tronic Article Surveillance (EAS) functionality.

Although special transponders that can be reused

are more expensive than transponders for one-time

use, using the same tags multiple times leads to signif-

icant cost savings. GW anticipates that each transpon-

der will circulate 8 to 10 times on average. There are

three reasons for this conservative estimate: (1) the

surface of the plastic tags wears out over time which

could have a negative effect on the perceived quality

of GW’s products when being presented on the shop

ﬂoor; (2) the functionality of the RFID transponders

may suffer over time; (3) transponders may get lost in

the supply chain.

The savings resulting from the repeated use of

transponders are diminished by the additional ex-

penses for shipping the transponders upstream. How-

ever, these expenses are comparably small because

the capacity of containers and trucks on their way

back to the production sites in Eastern Europe, Turkey

and the Far East are usually not utilized and therefore

relatively cheap.

Another advantage of closed-loop applications is

that object identiﬁcation does not require globally

unique identiﬁers such as Electronic Product Codes

(EPCs). Although GW is a member of the lead-

ing RFID industry consortium EPCglobal and also

pays fees for the use of European Article Numbers

(EANs), actually using the EPC numbering system

with full service would result in additional licensing

fees. Therefore the EPC is not used for now.

RFID IN THE SUPPLY CHAIN: HOW TO OBTAIN A POSITIVE ROI - The Case of Gerry Weber

EAS tags are currently attached to the garments

in the European distributions centres. By combining

EAS and RFID, no additional RFID tagging process

is required. Moreover, both EAS and RFID function-

ality can be added at the manufacturing sites where

labour is a lot cheaper than in the distribution centres.

The additional beneﬁts realized by the combi-

nation of cross-company, closed-loop and integrated

use of RFID resulted in a sufﬁciently positive ROI.

However, the additional requirements imposed by

cross-company closed-loop integrated use of RFID

transponders also lead to a number of critical chal-

lenges in terms of system and process design. These

will be discussed together with the envisaged system

design in more detail in the following section.

3.4 Design and Scope of the RFID

System

The design and development phase of GW’s RFID

project is almost completed by now. In collaboration

with their service provider IBM, they have speciﬁed

detailed use cases as well as hardware and network

infrastructure for the following processes:

• Virtual association at suppliers, consolidation

sites, distribution centres, and retail stores (UC1)

• Goods issue at suppliers, consolidation sites, dis-

tribution centres, and retail stores (UC2)

• Goods receipt at distribution centres and retail

stores (UC3)

• Localization of unassociated garments at distribu-

tion centres (UC4)

• Disassociation of items at distribution centres and

retail stores (UC5)

• Stock taking at retail stores (UC6)

All use cases have to comply with the above-average

complexity and diversity of GW’s supply chain pro-

cesses. For instance, they have to be designed both for

automatic and manual distribution centres, for hang-

ing and lying garments, for the described reusable and

also one-time use tags (possible requirement in the fu-

ture), and for different locations. Figure 3 provides an

overview of the information ﬂow within GW’s sup-

ply chain, the involved information systems and the

scope of the planned RFID system. Due to space lim-

itations we will concentrate on the two most critical

RFID processes. We will point out the challenges in-

volved in their implementation and how the system

architects plan to cope with them.

Order Receipt

RFID

System

Production

Picking Process

Goods Issue

Transportation

Goods Receipt

Consolidation

Goods Issue

Transportation

Goods Receipt

Quality Assurance

Storage

Picking Process

Goods Issue

Transportation

Goods Receipt

Storage

Sale

UC1

GW‘s

Merchandise

Information

System

UC2

UC3

UC1

UC2

UC3

UC1

UC4 UC5

UC3

UC2

UC3

UC6

UC5

Production order

Production progress

Supplier ASN

Long-haul TP ASN

Transportation progress

Received goods count

QA Information

Inventory levels

Picking information

LSP ASN

Inventory levels

POS data

Supplier ASN

Long-haul ASN

LSP ASN

Supplier

Long-haul

Transportation

Provider

(LTP)

Logistics

Service

Provider

(LSP)

Short-haul TP

GW Store

GW‘s SCM System

GW‘s ERP System

Figure 3: Information ﬂow and scope of the RFID system.

3.4.1 The Virtual Association Process

In order to reuse RFID transponders, garments have

to be associated both physically and virtually. Phys-

ical association means that a transponder gets physi-

cally attached to a garment; virtual association means

that by an appropriate database operation a transpon-

der ID is associated with a garment stock keeping unit

(SKU), the ID of the transportation unit it is trans-

ported in and the context data object (e.g. a produc-

tion order) it belongs to. Whereas the association

of transponder ID and SKU ID is a basic necessity

for connecting read events with the appropriate con-

text data stored in the IT backend, the association of

transponder ID with the IDs of transportation units

and context data objects is required to enable the ef-

fective tracking of the ﬂow of goods; in particular,

these associations can be used to identify tagged items

which have not been observed at a particular RFID

checkpoint although they should have been observed

there according to the schedule. In addition, GW

plans to validate the completeness of garment lots al-

ready during the association process: based on a real-

time matching of association data and available busi-

ness context data workers will be instructed to take

corrective action if necessary.

Both association types (physical and virtual) can

be conducted at several sites throughout the supply

chain (at large suppliers, consolidation sites, DCs) in

order to make sure that 100% of the garments which

leave the distribution centres are physically as well as

virtually associated (otherwise, the intended store use

ICEIS 2009 - International Conference on Enterprise Information Systems

100

cases such as stock taking could not be carried out).

To this end, the association status of every shipment,

i.e. ’untagged’, ’tagged’, ’tagged and virtually associ-

ated’, is captured and relayed so that each of the fol-

lowing supply chain partners can react accordingly.

The RFID reader devices which will be used at the

work stations responsible for conducting virtual asso-

ciation will be able to read both barcodes and the data

stored on RFID transponders. Each station will ac-

commodate a dedicated server which will temporarily

buffer RFID-related data such as the business context

data required in the association process and the results

of the association process. For hanging garments, a

handheld device will be utilized, whereas the virtual

association process for lying garments will be per-

formed on a specially designed RFID packing table

which enables ergonomic and fast processing. While

associating the garments, the items will be merged

into logistics units (either plastic bags or cartons).

Thereby, they will also be virtually linked to the num-

ber of the respective logistical unit.

Physical association is straight-forward since it is

similar to other standard labelling processes that are

already conducted along the supply chain (such as the

application of price labels). There are, however, a

number of speciﬁc challenges involved in the man-

agement of the EAS-RFID hardcase tags as well as in

the implementation of the virtual association process.

The RFID implementation strategy implies that

transponders are shipped to the large producers, the

consolidation sites and the DCs at the right time

and in a sufﬁcient quantity. GW expects that each

transponder is used two to three times per year de-

pending on the country it is shipped to, the transporta-

tion mode used, etc.. The calculation of the total num-

ber of required tags is not straight-forward since it has

to take buffer stock of tags into account.

Concerning the virtual association, the applica-

tion, among other things, has to be robust with respect

to the following exceptions:

• Non-readable RFID transponders

• Several RFID transponders in detection area

• Non-readable barcode

• Missing business context data

• Bypassing of predetermined processes by em-

ployees

Whereas the ﬁrst three issues can be solved pragmat-

ically (e. g. by replacing an RFID tag; shielding

or containment of the detection area; regeneration of

barcode label), the remaining two cannot be coped

with that easily: In case of missing business context

(such as purchase orders, ASN, etc.) immediate ac-

cess to the central RFID platform is required in order

to perform the planned quantity control. If the Inter-

net connection to the RFID system cannot be estab-

lished, the business logic has to be designed in a way

that conducting the virtual association without quan-

tity control is still feasible. Furthermore, the process

either has to be able to mitigate the faulty virtual as-

sociation of garments or at least detect it and notify

subsequent stages of the supply chain so that the as-

sociation process can be selectively repeated later on.

3.4.2 Goods Issue Process

When garments have been prepared for goods issue

at an RFID-enabled site, they have to be both physi-

cally and virtually associated as well as packaged in

plastic bags (hanging garments) or cartons (lying gar-

ments). The goods issue marks the last opportunity

to ensure the completeness of a shipment and the cor-

rect association of RFID transponders with garments.

In GW’s supply chain RFID hardware has to comply

with special environmental conditions. For instance,

the goods issue process can be conducted manually,

via suspension rails, with pallets or on conveyor belts.

In order to satisfy this requirement, different RFID

hardware solutions are necessary. Warehouse opera-

tors can for instance choose to use handheld devices

or a ﬁxed RFID gate. Another challenge regarding

the RFID-enabled goods issue process is to read the

data stored on many transponders at the same time:

since reading reliability can still not be guaranteed,

ways to make the process robust to below 100% read-

ing rates had to be found. To this end the allocation

of transponder IDs and logistical unit IDs is used: if

all items packaged in the same carton or pack can be

inferred from the database, reading one transponder

ID sufﬁces to conduct a full count at the goods issue.

Similar to the virtual association process de-

scribed above, GW will realize value added at the

goods issue by detecting potential deviations from the

distribution schedule. Warehouse operators will have

access to the RFID-based real-time loading status and

can be alerted if a shipment is incomplete. The RFID

system will be able to display to workers which items

or logistic units are missing in order to complete a

shipment.

3.5 Project Management

The design and implementation of the described

RFID application will be carried out by one main con-

tractor (IBM) and several subcontractors (hard- and

middleware suppliers) supervised by IBM. The ad-

vantage of this organizational model for GW is to

have one single point of contact for managing all of

RFID IN THE SUPPLY CHAIN: HOW TO OBTAIN A POSITIVE ROI - The Case of Gerry Weber

101

its RFID operations. IBM is responsible for the de-

sign, trial, rollout, and operation of the RFID system.

In order to minimize the ﬁnancial risk involved in

realizing the RFID application, GW decided to ad-

vance in four steps. After each of these steps the con-

tinuation of the project will be re-evaluated. In the

ﬁrst step, the RFID solution is designed, tested and

implemented in four distribution centres and two re-

tail stores. In the second step, the solution is rolled out

in all of GW’s retail stores. The international rollout

of the solution is initiated in the third step by equip-

ping two consolidation sites in China and Turkey with

RFID technology. In the ﬁnal step, all remaining con-

solidation sites in the Far East will be included which

will enable the maximum item-level visibility that can

be achieved using RFID.

4 CONCLUSIONS

This paper described the current state of Gerry We-

ber’s RFID project. GW is one of the ﬁrst companies

in the apparel industry that has decided to introduce

RFID on the item level. Based on the description of

the ﬂow of goods and information, our case study dis-

closed the beneﬁts GW expects from the introduction

of RFID. The case description gives insights into the

processes taking place along GW’s supply chain and

the data being exchanged between the supply chain

partners. GW has outsourced most of their logistics

operations but at the same time has to face the increas-

ing complexity of their supply chain. Therefore they

are interested in monitoring the ﬂow of goods more

closely in order to diagnose possible exceptions re-

motely. Besides typical RFID beneﬁts such as labour

cost and error reduction, RFID is expected to play a

key role in providing accurate monitoring data, which

will be used by sophisticated decision support sys-

tems.

The case shows how the ROI of RFID hardware

can be improved by closing the tag loop and ef-

ﬁciently combining RFID with extant technologies

such as EAS. Two RFID use cases that play a cru-

cial role in the implementation of GW’s RFID strat-

egy were analyzed: the virtual association and the

goods issue process. The corresponding advantages

and challenges were outlined. The approach of using

RFID in a closed loop and integrating it with extant

EAS processes is not necessarily restricted to the ap-

parel industry. It could also be applied to other retail

products of relatively high sales value (e.g. consumer

electronics). In our opinion, the approach of increas-

ing the ROI of item-level tagging outlined in this pa-

per has the potential to foster the diffusion of large-

scale item-level RFID applications. If more RFID

transponders are sold their price will eventually drop

due to the economies of scale realized in their pro-

duction. This in turn will eventually also make open

loop applications economically viable. Although GW

will start to use RFID transponders in a closed loop,

the system architecture that is currently being imple-

mented can easily be expanded to an open-loop appli-

cation: when GW’s wholesale customers start calling

for RFID tagging, GW will be ready to serve their re-

quest.

REFERENCES

Chappell, G., Durdan, D., Gilbert, G., Ginsburg, L., Smith,

J., and Tobolski, J. (2003). Auto-ID in the Box: The

Value of Auto-ID Technology in Retail Stores. Tech-

nical Report, Auto-ID Lab MIT.

Chawathe, S. S., Krishnamurthy, V., Ramachandran, S., and

Sarma, S. (2004). Managing RFID Data. In Proceed-

ings of the Thirtieth International Conference on Very

Large Data Bases, pages 1189–1195.

Gaukler, G. M. and Seifert, R. W. (2007). Trends in Sup-

ply Chain Design and Management: Technologies and

Methodologies, chapter Applications of RFID in Sup-

ply Chains. Springer London Ltd.

unther, O. and Spiekermann, S. (2005). RFID and the

Perception of Control: The Consumer’s View. Com-

munications of the ACM, 48(9):73–76.

IDTechEx (2007). RFID Forecasts, Players and Opportu-

nities 2007-2017. http://www.idtechex.com/forecasts.

Last visited: 11/15/2008.

Juels, A. (2006). RFID Security and Privacy: a Research

Survey. IEEE Journal on Selected Areas in Commu-

nications, 24(2):381–394.

Lee, H. and

Ozer, O. (2007). Unlocking the Value of RFID.

Production and Operations Management, 16(1):40–

64.

Loebbecke, C. and Huyskens, C. (2007). Towards Standard-

izing Success: RFID in Fashion Retailing. In Proceed-

ings of the 20th Bled Conference.

Ngai, E. W. T., Moon, K. K., Riggins, F. J., and Yi,

C. Y. (2008). RFID Research: An Academic Litera-

ture Review (1995-2005) and Future Research Direc-

tions. International Journal of Production Economics,

112(2):510–520.

Schmitt, P. and Michahelles, F. (2008). Economic Impact

of RFID. Report, BRIDGE Project.

Strassner, M. (2005). RFID im Supply Chain Management.

Deutscher Universit

ats-Verlag, Wiesbaden, Germany.

Voss, C., Tsikriktsis, N., and Frohlich, M. (2002). Case Re-

search in Operations Management. Int. Jounal of Op-

erations & Production Management, 22(2):195–219.

Wu, N. C., Nystrom, M. A., Lin, T. R., and Yu, H. C. (2006).

Challenges to Global RFID Adoption. Technovation,

26(12):1317–1323.

ICEIS 2009 - International Conference on Enterprise Information Systems

102