examining rfid applications in supply chain management

COMMUNICATIONS OF THE ACM July 2007/Vol. 50, No. 7 93

Radio Frequency Identification(RFID) is an emerging tech-nology recently in the news as aresult of large organizations,most notably Wal-Mart and

the U.S. Department of Defense, requiringshipment of goods identifiable by RFIDtags. In Europe, the Metro Group, the fifth-largest retailer in the world, has teamed withSAP, Intel, and IBM to form the FutureStore Initiative (www.future-store.org). Thisgroup is using a combination of differenttechnologies in a test store to learn whetheran integrated information system can lowercost by improving inventory management,consumer information gathering, andcheckout procedures. As of March 2006,

Technologyinfrastructure, business process, and managerialissues must beaddressed by ITpractitioners as they adapt to thebusiness changesassociated with thediffusion of RFIDtechnology in thesupply chain.

I L L U S T R A T I O N B Y P E T E R H O E Y

EXAMINING RFID APPLICATIONS IN SUPPLY

CHAIN MANAGEMENT

BY FRED NIEDERMAN,RICHARD G. MATHIEU,ROGER MORLEY, AND

IK-WHAN KWON

i

94 July 2007/Vol. 50, No. 7 COMMUNICATIONS OF THE ACM

Wal-Mart reported establishing RFID programs withmore than three times the 100 major suppliers initiallytargeted in 2005 [10]. Extensive RFID programs havealso been established by the U.S. Department ofDefense.

Operationally, RFID allows data froma tag attached to a pallet, case, orindividual product to be captured by a reader device. Functionally, thisdata can be used to identify all ofthese items passing the reader’s loca-

tion at a point in time. This technology provides ameans of tracking items from supplier through the dis-tribution network to the point of consumption.Recent advances have lowered the cost of tags suchthat passive (not battery powered) tags are beginningto approach commercially viable levels.

The goal of RFID technology applied to theretail supply chain is to streamline inventory man-agement by providing views of product shipmentsand inventory levels at unprecedented levels ofdetail [11]. By providing precise data on productlocation, product characteristics, and productinventory levels, RFID promises to eliminate man-ual inventory counting, warehouse mispicking, andorder numbering mistakes. The ultimate goal is thereduction of inventory and increased product salesin the retail supply chain.

From an information systems perspective, RFIDcan be viewed as another technology for capturingsource data. But the characteristics of RFID in thesupply chain context—that it can be read unat-tended, in groupings of various sizes, through somematerials and with a level of flexibility regardingorientation and positioning between tag andreader—creates the possibility that RFID use willbecome transformational. Rather than simply sub-stituting for existing technologies such as bar codeswithin current processes, it holds the promise ofprocess innovations that create orders of magnitudeimprovements in supply chain business models.Achieving such benefits will require integratingRFID technology into supply chain operations.

Currently, from a supply chain perspective, sev-eral major technical challenges exist for firms wish-ing to initiate RFID programs. These challengesrevolve around the mechanics of reading RFIDtags in the context of “shop floor” realities. Some ofthese challenges are due to difficulties with deflec-tion and refraction from materials such as metal,glass, and liquids; difficulties with angles andreception; and difficulties attaining the levels ofvolume and speed required to support industrial

supply chain information requirements. As thesetechnical challenges find engineered solutions,organizations will have a new longer-range set ofchallenges to create significant economic value.These will revolve around justifying initial invest-ment in the technology that would appear toinvolve tangible costs but less tangible benefits;developing the information systems solutions thatwould integrate data derived from RFID into busi-ness processes; and developing the “feedback-loops” such that new sources of informationgenerate more efficient and effective businessprocesses and decision making.

Here, we identify the range of issues IT practi-tioners are likely to confront in fully implementingRFID technology. These issues and ideas are dis-cussed within a data life cycle framework to helporganize the wide range of topics into related sets.Because this technology was relatively new and theauthors knew of no examples of fully instantiatedRFID-integrated information systems in supplychain management, the framework is presented as ahelpful viewpoint rather than as a validated optimalapproach.

The key issues organizations will face in derivingmaximum benefit from RFID technology as itmatures are previewed here. Most prior researchfound in the ACM digital library pertaining toRFID application development falls into two cate-gories. Either it demonstrates the potential offuturistic applications such as the use of RFID inubiquitous computing for outdoor educationalexperiences (see [12]) or focuses on issues of pri-vacy and ethical use of RFID information, (see [1,4]). Although both are important topics, they donot need to be discussed again here; instead, wefocus specifically on RFID use in supply chainmanagement.

RFID BACKGROUND

RFID technology is derived from World War II-eratechniques to allow aircraft to identify themselvesto other friendly aircraft and commanders on theground. Before the invention of the transistor andthe evolution of microelectronics, the technologywas large, heavy, and consumed massive amountsof power. In the 1970s, Sandia National Laborato-ries began incubating commercial applications andlaunching businesses. One of these was the AmtechCorporation, which commercialized animal tagging(still used today to identify pets), and the motorvehicle account tag used on some toll roads startingin the 1980s. At about the same time, industrialapplications in manufacturing automation, ware-

house automation, and assettracking were deployed. In the1992–1994 timeframe, theNorth American rail industry

deployed the Amtech technology for tracking railequipment, tagging over 30,000 locomotives and1.2 million rail cars and deploying reader devices atkey control points along the railroad right-of-way.

From 1998 through 2003, leadership of retailinitiatives has been centered at MIT where in 1999the AutoID center was established. In late 2003,the AutoID Center at MIT officially closed andtransferred its intellectual prop-erty to EPC Global. (It is impor-tant to note that EPC Globalstandards relate only to thedeployment of RFID into theretail supply chain.) The pres-sure is intense within enterprisesfacing a mandate to comply withthe Electronic Product Code(EPC) Global standard, a jointventure of the Uniform CodeCouncil and European ArticleNumbering (EAN) Interna-tional, the body that administersthe EAN standards worldwide.(See www.epcglobalinc.org for acomprehensive description of theEPC Global mission, strategy,and objectives.)

EPC Global standards areattracting significant attentionbecause of the Wal-Mart man-date that its top suppliers beginadopting them. Also, severalother large retailers have joined the initiative, andthe U.S. Department of Defense has issued a policydirecting its suppliers of similar product to complyas well. The standard is also important in establish-ing the potential for use of RFID information across

organizations in a standardized format. Today assuppliers struggle to implement RFID solutions,they are just beginning to grapple with the issue ofhow to integrate new RFID with existing enterprisedata [11]. Considering the investment in themechanics of using RFID is significant and tangi-ble, the leverage gained from that information beingable to streamline not only internal businessprocesses but to create efficiencies between businesspartners is also important in shaping the discussionof ultimate costs and benefits of diffusion of thistechnology.

RFID DATA THROUGH ITS LIFE CYCLE

In order to examine the challenges and opportuni-ties for IT professionals in RFID implementation,it is useful to follow the data generated by readingan RFID tag and to consider the journey of thatdata as it travels through the data life cycle. Under-standing the data life cycle is important to under-standing the nature of data. (Note also that wepresent something of an ideal view of the journeyof particular data from beginning to end of the lifecycle while recognizing that at each identified stagetechnical and organizational issues must be

resolved before such a journey can become routineand cost effective.) Redman proposed a data lifecycle that includes two cycles: the data acquisitioncycle and the data usage cycle [5]. Recently, therehas been a renewed interest in the data life cycle asIT vendors are developing policy-driven automatedstorage architectures that are able to manage datafrom beginning to end [2]. For the purposes of thisarticle, we view the data life cycle in a series of sixstages:


Figure 1. Data Life Cycle stages for RFID-generateddata in supply chain management.

Niederman fig 1 (7/07)

Stage 2:Integratingwithin theenterprise

Organization Application Environment

Stage 3:Integratingacrossorganizations

Stage 4:Data warehousingand analyticalprocessing

Stage 5 and 6:Archiving,deletion,and disposal

Data Life Cycle FrameworkPhysical Supply Chain

Tags

Stage 1:Source DataAcquisition

Figure 1. Data life cyclestages for RFID-generateddata in supply chain management.

Figure 2. Typical applications and data within the supply chain data lifecycle.

Niederman fig 2 (7/07)

Stage 4: Warehousing and analytical processing

Supply Chain Enterprise

OrderManagement

TransportationManagement

WarehouseManagement

Supplier Rel.Management

ProductionManagement

Customer Rel.Management

EPC GlobalIS*

EDI**

E-Mail

OperationalData Store

Enterprise DataWarehouse

IS = Information ServiceEDI = Electronic Data Interchange

Decision Support,Quantitative Analysis,

Modeling,

Extraction, summarization, organization

OperationalData Store

OperationalAnalysis

Controller

Managers

ExecutivesWorkers

Reader

OperationalAnalysis

OperationalAnalysis

Tags AntennaPhysical Supply Chain

Figure 2. Typical appli-cations and data within

the supply chain datalife cycle.

• Source data acquisition (read-ing RFID data and integrat-ing with other source data);

• Integrating source data withenterprise transactions sys-tems;

• Integrating data across organi-zations;

• Data warehousing and ana-lytic processing;

• Data archiving (backup andreplication); and

• Data deletion and disposal.

These stages are illustrated inthe supply chain managementcontext by Figures 1 and 2.Each conceptual stage presentschallenges and opportunitiesfor the use of information toenhance supply chain activitieswithin and among organiza-tions. Though this sequence islogical in terms of aggregation,it should not be taken as literallydescribing the sequence ofactions in actual usage. Forexample, in some scenariosRFID data might be used along with interorganiza-tional data concurrently in performance of a partic-ular business task. Nevertheless, the stage frameworkprovides a logical way of decomposing the poten-tially highly complex and recursive nature of usingRFID data.

It is a reality of RFID deployment that, to be suc-cessful, it involves changes in organization, businessprocesses, and application systems. Therefore, analy-sis of the data life cycle involves examination alongthese dimensions. Dividing application systems intoinfrastructure and logic, we present our analysis insubdivisions: technology infrastructure; applicationlogic; business process; and management. Detailsregarding RFID at each stage on each dimension arepresented in Tables 1 and 2.

Source Data Acquisition. An IT perspective onRFID begins with initial handling of the data gen-erated by the RFID reader. All readers and con-trollers must be able to distribute data to theoperational applications in distant locations. Thisimplies that all devices capturing data must be con-nected either through wired or wireless networks.The key will be to ensure all physical elements ofthe system have sufficient capacity so that data isreceived, processed, and moved without bottle-

necks that inhibit the organization’s ability tomaintain the speed of the business activity wherethe RFID is deployed.

Once elemental data (originatingfrom RFID) has been received andvalidated, it is likely that additionalinformation will be added includ-ing time, temperature, location,and similar indicators. This exten-

sion of data surrounding RFID messages is beingcalled “sensor-based computing” by practitioners atOracle [7]. For the item being identified, it is self-evident to human workers where the tag is beingread, but for downstream use of this data, time andlocation information needs to be added.

Programming logic for the disposition ofexpected and unexpected RFID-generated data willneed to occur at each stage in a supply chain bothinternally to an organization and among organiza-tions. Accurate timing of such messages will also becritical. As a simple example, timing calculationmust account for time zone changes so that itemsdon’t appear to arrive before they are sent. Further,the meaning of each tag reading may need to be clar-ified with additional information. For example, doesa message indicate shipment of goods, receipt ofgoods, movement into or out of inventory, or tem-porary removal for inspection by a customer versuspermanent removal for sale or disposal?

From an application logic perspective, muchwork needs to be done handling too much, too lit-


Table I: Issues in implementing RFID Applications –Data Life Cycle (Stages 1 and 2) vs. System Layers

Niederman table 1 (7/07)

DATA LIFE CYCLE

1. Source data acquisition (reading data and integrating with other source data)

2. Integrating within the enterprise (integration with enterprise systems and supply chain applications)

TechnologyInfrastructure

Performance, speed, and capacity

Optimal processor, data storage, and network bandwidth elements

Accuracy


Timing

Reliability


Middleware

Application Logic

Business logic – capturing the range of valid actions

Missing data

Unwanted data

Information overload and filtering (too much wanted data)

Generate metrics

Integration with business units’ information

Cross references between RFID information and other enterprise system keys

Normalization and concentration of information by logical entity

Business Process

Integration with physical processes/ logistics/industrial engineering

Identifying elemental logical operations – definition of business events

Timing so that reading occurs at correct point(s)

Discovery of business logic exceptions

Application design/development

Covering basic operations

Imaginative new programs

Potential performance implications of enterprise update on operational process

Management

Responsibility for investigating exceptions/ recognizing new problems

Responsibility for standard variations in process/organizational learning

Metrics/evaluation(error rates,throughput)

Availability of programmers for maintaining manufacturing system

Responsibility for investigating exceptions/ recognizing new problems

Responsibility for standard variations in process/organizational learning

Company specific architecture knowledge

Metrics/evaluation

Table 1. Issues in implementing RFID

applications: Data lifecycle (stages 1 and 2)

vs. system layers.

tle, and unexpected data. For example, the tag willrespond repeatedly to a reader’s request for data aslong as the two are in proximity. In many cases,only the ultimate arrival and departure of the item(change of state) will be useful to downstreamapplications.

To the extent that semantics of the data areaccounted for, some strategies found to be useful inother high throughput, online tasks, such as thedistribution of financial data, can be employedwith RFID. Palmer identifies seven such strategiesincluding process data close to its source, clustervarious pieces of data into logical events, utilizedata concentrators, cache contextual information,federate data locations for more efficient distribu-tion of data among sites, continuously filter dataevents, and automate exception handling [3].While it certainly doesn’t pay to store redundant ornon-useful data, the planning for these strategies,particularly filtering, must be undertaken carefullyso that elements necessary for later analysis are notremoved permanently.

From a management perspective, it will beimportant to build systems that are robust enoughthat such data will be collected and integrated evenwhen pressured human operators are more con-cerned with “getting the job done” than ensuringaccurate acquisition of information needed only atlater stages in the business process. Capitalizing onimaginative supply chain applications may requireextremely close to 100% accuracy in upstreamoperations and millisecond (or faster) time frames

for downstream analysis ofincoming orders.

Integration with EnterpriseSystems. Organizations willface different sets of issues inintegrating RFID with existingsystems depending on the typeof systems they already have. Weanticipate four basic scenarios:organizations with relatively adhoc business processes will gen-erate flat files from RFID andsensor data and provide this datafor manipulation by existingapplications; organizations willbuild new relational databases,systems, and processes (manualand automated) to interact withexisting enterprise systems; orga-nizations will work with vendorsto build capabilities into theenterprise systems such that theyhandle the data originating fromthe RFID sources; or organiza-tions with custom legacy sys-tems will take this opportunityto convert to enterprise systemsand simultaneously address sys-

tem integration and utilization of RFID capabili-ties. Each of these scenarios presents opportunitiesand risks. Similarly, the amount of investment inand optimization of bar code scanning technologywill affect the distribution of costs and benefitsfrom shifting to RFID. In the long run, however, asRFID becomes pervasive and is integrated withother sensor information, opportunities may ariseto use this richer information to create innovativenew business models.

Vendors such as SAP, Oracle, Sun, Peoplesoft,IBM, and Microsoft are currently deploying vari-ous middleware approaches toward the integrationof RFID with their existing product offerings. Forexample, SAP is reported to have developed a mid-dleware layer, named the Auto-ID Infrastructure,that routes data from readers to applications(including multiple communication and sensingdevices such as RFID readers and printers, Blue-tooth devices, embedded systems, and bar-codedevices) and triggers appropriate events using arules engine [6]. This approach is likely to empha-size reformatting incoming data for use with legacysystems of various types.

From a technology infrastructure perspective,accuracy and planning for contingencies will be


Table 2. Issues in implementing RFID Applications –Data Life Cycle (Stages 3-5) vs. System Layer.

Niederman table 2 (7/07)

DATA LIFE CYCLE

3. Integrating across organizations (communication and coordination with EPC Global, EDI and/or email applications)

4. Data warehousing and analytic processing

5. Archive (backup and replication)

6. Deletion and Disposal

TechnologyInfrastructure


Timing and Reliability


Architecture and Security

Massive amounts of data and potential capacity issue

DSS, data mining,visualization,knowledge management tools

Telecom -- middleware

Automated storage solutions

Application Logic

HW and SW engineering

Standard Web tools

Transaction information standards

Translation of firm specific information to commonly accessible languages

Aggregating the data

Data cleansing

Policies for when data no longer needed

Business Process

Imaginative new programs / value added services

Potential performance implications of partner system response on operational process

Add decision systems at multiple new levels

Knowledge creation (Smart shelves,in-store customerbehaviors, end-to-end supply chain analysis

Processes for deleting and disposing of data

Management

Deciding what to share and what not to share (and with whom)

Negotiation with partners

Ethics, liability, public perceptions/ privacy

Metrics/evaluation

Emphasis on knowledge management (creation,storage, transfer)

Allow new layers of drilldown in operational events

Metrics/evaluation

Metrics/evaluation

Legal issues

Records management

Table 2. Issues in implementing RFIDapplications: Data lifecycle (stages 3 through5) vs. system layers.

critical. Downstream operations cannot be stoppedwithout undue cost if data from upstream opera-tions has been lost, misdirected, or become inaccu-rate. This also raises issues of capacity and assuringthat storage, processing, and network capacities aresufficient for handling with the vast quantities ofdata generated by RFID and the processes withwhich they may interact.

Akey issue in implementing such asystem will be developing appro-priate business rules, particularlyfor handling new exceptions thatarise from having the increasedamount of data generated from

RFID use. It can be expected that in many casesexplicit rules will need to be created for program-ming business activities where informal methodsare currently in use.

From an application logic perspective, it will becritical to normalize data. This is significant fordealing with dependencies and assignment ofattributes to proper entities and in the managerialsense of ensuring that terms are unambiguouslyand appropriately defined for business needs.

While it addresses ways to automate data inputinto current business processes, in the long run, it islikely that the presence of RFID will present oppor-tunities to develop innovative business models andto reengineer existing supply chain processes.Examples can be found for many common supplychain activities. The ability to determine time andlocation of reading from unique items will provideopportunities to identify bottlenecks. When com-bined with value of shortened cycle time, this infor-mation can also potentially influence deliverypriorities where resources are constrained. The abil-ity to determine where a particular object is and hasbeen can solve many supply chain problems includ-ing the ability to shorten the time to effect productrecalls and removal of outdated products, programsto decrease product counterfeiting, decrease occur-rences of out-of-stocks, and reduction of shrinkage,and diversion of products. The ability to locateinventory can shorten the time for finding a prod-uct as it moves between production floor, ware-house, and showroom floor. Use of RFID-generatedinformation may allow the clustering of small pack-ages from various sources heading toward the samedestination early in the distribution channel or auto-mate dynamic rerouting as information regardingtravel conditions or package priorities change. Onerelated challenge will be dealing with the dynamicaspects of systems change—designing the structure

and content of RFID messages (integrated withother sensor data) while simultaneously reconsider-ing business processes and the IT supporting them.

From a management perspective, the key issuewill be balancing innovation and the effort toachieve competitive advantage against the risks andcosts associated with groundbreaking activities. Thecosts of handling the vast amount of data generatedby RFID will be largely tangible and occur early inthe data life cycle while the benefits will depend onthe quality of data collection, will require addedeffort and investment, remain largely intangible,and be recovered later in the business cycle. More-over, those firms aggressively committed to analysisand reorganization of business processes will havethe risk that improvements will either not be foundor be too difficult to implement.

Integrating Across Organizations. Much of therationale for adopting the EPC Global architecturein the retail supply chain is to facilitate sharing ofinformation among supply chain participants suchas supplier, manufacturer, shipper, and customer inorder to provide near-real-time inventory visibility.The value created by this process is reduced costand improved competitiveness of the cooperatingenterprises. Cost reductions will result from reduc-ing the amount of inventory and increasing logis-tics efficiency in the supply chain. These includethe cost of carrying the asset, direct network costssuch as warehousing and transportation, and obso-lescence. Better visibility will also reduce the inci-dence of expensive interventions to avoidundesirable shortages. Improved competitivenesswill result from higher customer satisfaction andmarket responsiveness, and from the ability to dif-ferentiate commodity products with creative logis-tics and marketing programs.

The EPC Global architecture originated at theAuto-ID center at MIT. Key principles underlyingthis architecture included minimizing functional-ity on the chip (and thereby lowering the cost); aproduct identification scheme that involved theenterprise, the product, and the item; using theInternet to store and share information about theitem; and open standards. The core component ofthis architecture is the product identifier, nowcalled the Electronic Product Code (EPC). TheEPC is a set of format standards for encoding prod-uct identification data on the tag. In the originalconcept, the EPC had four components, header,EPC manager (the enterprise,) object class (theproduct,) and serial number (the item). The origi-nal standard is being replaced by a more complexstructure to ease the transition from existing stan-


dards. The other components of the architectureinclude:

• Information Service (IS). A repository of theenterprise’s product information, indexed byEPC and accessible via the Internet;

• Object Name Service (ONS). A global registryof EPC information; and

• Product Markup Language (PML). A markuplanguage specific to EPC data storage andretrieval.

Plans for an ONS that mirrors Internetdomain name addressing present a nearlyoverwhelming set of opportunities andchallenges. For example, infrastructuralchallenges of integrating across organiza-tions will include creating and supporting

applications that enable other enterprises to accesstheir information systems. This will likely require thedevelopment of additional security capabilities.

The IS can be viewed essentially as a new appli-cation. It must be designed and implemented eitherby the participating enterprise or by a third-partyservice. In addition to the core function of the IS,application logic will be required to support a mul-tidimensional security model, and to maintainONS registration. Full compliance with the EPCGlobal architecture will also include PML support,and this may impact the logic of existing internalapplications.

Business process impacts include the coordina-tion of PML dialects with trading partners and theadministration of the multidimensional securitymodel mentioned previously. Two decades of expe-rience with EDI have demonstrated that even witha relatively rigid format standard for interenterprisemessage management, trading partners must coor-dinate their individual implementation of thosestandards. The use of PML will create similarrequirements such as developing understandingswith partners regarding the definition of exchangedinformation, the timing of messages, and the inves-tigation and resolution of apparent exceptionsresulting from operational mistakes, data entryerrors, and true business misunderstandings. Sincethe EPC Global architecture provides for the accessof one enterprise’s data by another enterprise, eachparty will need a process to administer the securityof its information. At a minimum, the identifica-tion, authentication, and authority of each partici-pant in the relationship must be specified withrespect to what functionality may be used and whatdata may be accessed.

Finally, the management impacts of EPC Globalcompliance will include the need to set policy withrespect to the trading partner relationships andnegotiating with trading partner management.There will also be a need to manage cultural expec-tations at all levels as trading communities movefrom a world of point-to-point relationships tomassive collaboration. The current initiatives ofWal-Mart and others constitute a limited steptoward the full EPC Global vision. There are nocurrent plans to tag and identify individual inven-tory items, only transportation units (pallets, car-tons, cases, and other packaging). Until inventory isidentified at the item level, the benefits of andrequirements for the IS, ONS and PML are limited.

Data Warehousing/Analytical Processing. Tothis point, expected benefits from the addition ofRFID-generated data have been discussed in termsof efficiencies in supply chain operations and newapplications that could provide new services or pro-vide additional efficiencies. However, the separa-tion of physical products and information aboutthem may have its most dramatic influence in sup-porting decision making and other managerialactivities. The RFID data warehouse must main-tain a significant amount of data for decision mak-ing. Historical and current data is required fromsupply chain partners and from various functionalareas within the firm in order to support decisionmaking in regard to planning, sourcing, produc-tion, and product delivery. Supply chains aredynamic in nature. In a supply chain environmentit may be desirable to learn from an archived his-tory of temporal data that often contains someinformation that is less than optimal. In particular,these environments are typically characterized byvariable changes in product demand, supply levels,product attributes, machine characteristics, andproduction plans.

Though there is cost associated with storing thisdata, there is potential value as a raw material forknowledge creation, decision support, and datamining. The nature of this data is threefold: itemscan be identified at finer levels of individuality;there are many more business events for an item(events can be disaggregated into parts); and beingserialized, individual steps in a sequence have anidentity. This opens whole new fields for analysis.Not just 20 items leaving a warehouse and 10 arriv-ing at another location, but which items arrived, bywhat route, and how long they were stored.

From a technology infrastructure perspective,issues will involve selecting and implementing theright set of decision support systems and knowl-


edge support tools as well as organizing the data tomaximize the trade-off between capturing all pos-sible data and retaining a size that is manageablefor ad hoc as well as programmed queries. Applica-tion logic will be important in assuring the cleans-ing and structuring of data as it enters the datawarehouse and in the design of standard manage-ment reporting processes. Although the motivationfor using RFID in large part revolves aroundstreamlining operations, Shapiro estimates that80% of the data in a transactional database thatsupports supply chain management is irrelevant todecision making, and that data aggregations andother analyses are needed to transform the other20% into useful information [8]. The value of thedata warehouse and the collection of RFID processwill primarily be in the discovery of new relation-ships and opportunities for process redesign.Specifically within the supply chain context, thisinformation can aid in logistics network design,supply chain planning, and vehicle routing andscheduling [9]. It can potentially also provide newbusiness insights. IT managers will need to con-sider trade-offs in developing standalone RFID-oriented data warehouses versus integratingRFID-oriented data with other organizationalinformation from marketing, finance, andaccounting areas. From a management perspective,RFID information provides the potential to movecompanies toward the integration of marketingand operations and financial information. The dis-solution of barriers in these areas will have signifi-cant implications for the way organizationalstructures are drawn, the range and type of skillsneeded by managers, and the approach towardbusiness problems and service of customers thatorganizations can take.

Archive (Backup and Replication). Archivingdata is a fairly well understood process throughoutorganizations. By their very nature, RFID applica-tions will be generating large amounts of data. Forsome organizations, RFID data will shift their dataarchiving function to another level of complexityas it handles the significantly increased volume. ITmanagers will seek to shrink storage requirementsthrough strategic use of summarization. Strategiclogical and physical data organization will beneeded to balance complete data storage with effi-cient retrieval. Will data be more effectively storedbased on real-world entities, based on time andsequence, or some other categorization method?This is likely to be a company, perhaps even appli-cation-specific decision. Otherwise, we would notexpect much difference in issues of archiving from

other transaction systems that are critical for busi-ness operations, received in high volume, and ofhigh short- and long-term value.

Deletion and Disposal. The automated deletionand disposal of data records could be a significantissue in an RFID-saturated IT environment.Because of the volume of data, it will almost cer-tainly be necessary to use automated storage tech-nologies based, in all likelihood, on rules rangingfrom simple time-oriented to complex subject- orsource-oriented differentiations. These rules will becritical for maximizing the value of historic dataand meeting legal obligations without incurringdisadvantageous data storage costs. Organizationswill need to perform risk assessment to considerissues such as: How likely is it that we will needthis data again? What are the consequences of nothaving this data? What is the cost of keeping thedata? And does the data have potential futureresearch value of a currently unknown type?

CONCLUSION

We have used a data life-cycle framework to dis-cover and illuminate key issues that will face ITpractitioners as they approach business changesresulting from the diffusion of RFID technology.We have considered these issues along the dimen-sions of technology infrastructure, applicationlogic, business process, and managerial concerns.In conducting this analysis, we also noted someissues that persist across the life cycle, particularlyin the case of scalability and performance. How-ever, it is important to note that at each stage in thedata life cycle, managers will need metrics relatedto both business and technical performance. Thesemetrics will vary with the stage but in all casesshould relate to the overall value created from theuse of RFID technology as well as the efficientoperation of information services at each stage.Additionally, security issues abound. Though theyare probably similar to other e-commerce securityissues, the sheer scale may present new issues thatmust be addressed at a system level. From the per-spective of an organizational level, part of the busi-ness process management will be deciding on andimplementing various levels of access for differentstakeholders under dynamically changing circum-stances. Ethical issues also are present as the poten-tial for adding customer information and profilingbecomes attached to the creation and distributionof products. These may be similar to the potentialfor privacy violation with the creation and distrib-ution of intangible goods such as financial prod-ucts (stocks, bank deposits, credit card records) and


entertainment products (music and movies).Although there is no guarantee that RFID tech-

nology will spread in usage throughout the econ-omy, the potential benefits from continuing toreplace manual labor associated with the distribu-tion and storage of physical goods suggests astrong future for this technology. The immediateimpact of RFID technology will probably be expe-rienced primarily in the warehouses, loadingdocks, and transportation assets, and in the longrun full utilization of RFID technology willrequire significant business process redesign andassociated development of new and retooled IT.Clearly, RFID holds the promise of significantlyaffecting the handling of physical productsthroughout the supply chain. However, issues ofsecurity and privacy must be adequately addressedto balance market effectiveness with fair practicesfor that promise to be realized.

Although we did not approach thistopic from the perspective of theissues that confront the particularIT leader in undertaking an end-to-end RFID project, we recog-nize that such a perspective would

highlight additional issues and difficulties that ITleaders would need to master for successful imple-mentation. For example, at each stage IT leaderswill confront make/buy trade-offs considering dif-ferences in cost, timing, functionality, long-termsupport, and project risk in selecting packages ver-sus building new applications. We would anticipatethat some organizations will likely move slowlyapplying RFID technology only as required bylarger business partners or not adopting the tech-nology at all. Others, however, are likely to moveaggressively adding RFID capabilities in their oper-ations and develop new applications based on theopportunities presented.

We also did not directly address the topic ofwhat discipline within an organization would bemost effective at having responsibility for the intro-duction of RFID and related capabilities. An infor-mal poll of local supply chain managers suggested apartnership of supply chain and IT specialists as apromising model. It isn’t clear if organizations willbe better served by housing responsibility for RFIDtechnology in supply chain or IT departments(with consultation and inclusion of members fromboth areas) or if setting up a new organizationalcomponent with members drawn from each areawould be more effective. This may depend on thesize of the organization and the scope of its com-

mitment to utilization of RFID information. It canbe anticipated that in reality organizations imple-menting RFID will utilize a variety of governancestructures and future research would be needed todetermine under what circumstances each is mosteffective.

It is our hope that having identified and cata-loged some of the expected IT challenges andopportunities associated with the diffusion ofRFID technology, IT managers will be able toanticipate issues and measure the risks associatedwith its implementation more knowledgeably.

References1. McGinity, M. RFID: Is this game of tag fair play? Commun. ACM 47,

1 (Jan. 2004), 15–18.2. Mearian, L. The data life cycle. Computerworld (Nov. 17, 2003),

38–40.3. Palmer, M. Seven principles of effective RFID data management

(part 1 and 2). Enterprise Systems News; www.esj.com/news/arti-cle.asp?EditorialsID=1067.

4. Pottie, G.J. Privacy in the global e-village. Commun. ACM 47, 2 (Feb.2004), 21–23.

5. Redman, T.C. Data Quality for the Information Age. Artech House,Norwood, MA, 1996.

6. SAP launches RFID package. RFID Journal (Jan. 15, 2004). 7. Seeley, R. RFID and data: Here’s what’s next. Application Develop-

ment Trends (July 2004); www.adtmag.com/article.asp?id=9666.8. Shapiro, J.F. Modeling the Supply Chain. Duxbury, Pacific Grove,

CA, 2001.9. Simchi-Levi, D. Kaminsky, P. and Simchi-Levi, E. Designing and

Managing the Supply Chain. McGraw-Hill, Boston, MA, 2000.10. Songini, M. Wal-Mart offers RFID update. Computerworld (Mar. 13,

2006); www.computerworld.com/mobiletopics/mobile/story/0,10801,109418,00.html.

11. Wailgum, T. Tag, you’re late. CIO Magazine (Nov. 15, 2004),50–56.

12. Wilde, D., Harris, E., Rogers, Y., and Randell, C. The periscope:Supporting a computer-enhanced field trip for children. PersonalUbiquitous Computing 7 (2003).

Fred Niederman ([email protected]) is the Shaughnessy Professorof MIS in the Department of Decision Sciences and MIS at Saint LouisUniversity, MO. Richard G. Mathieu ([email protected]) is the chair and aprofessor of CIS in the Department of Computer Information Systemsand Management Science at James Madison University in Harrison-burg, VA. Roger Morley ([email protected]) is the vice presidentfor supply chain consulting at Protean, Inc. in Pacific, MO. Ik-Whan Kwon ([email protected]) is a professor of Decision Sciences and MIS in the Department of Decision Sciences and MISand the director of the Consortium for Supply Chain ManagementStudies at Saint Louis University, MO.

Permission to make digital or hard copies of all or part of this work for personal or class-room use is granted without fee provided that copies are not made or distributed forprofit or commercial advantage and that copies bear this notice and the full citation onthe first page. To copy otherwise, to republish, to post on servers or to redistribute tolists, requires prior specific permission and/or a fee.

© 2007 ACM 0001-0782/07/0700 $5.00

c


Documents

examining rfid applications in supply chain management