Gaurav Kapoor 1 , Wei Zhou 2,3 and Selwyn Piramuthu 1,3 1 University of Florida, U.S.A.; 2 Information Systems and Technologies ESCP Europe, France; 3 RFID European Lab @ESCP Europe Correspondence: Selwyn Piramuthu, Information Systems and Operations Management, University of Florida, Gainesville, FL 32611-7169, U.S.A. Tel: 1 (352) 392 8882; E-mail: selwyn@ufl.edu Received: 25 August 2009 Accepted: 6 October 2009 Abstract Radio-Frequency Identification (RFID) tags are gaining widespread popularity throughout the supply chain from raw material acquisition, manufacturing, transportation, warehousing, retailing to the ultimate consumers. A majority of extant literature in this area explore the beneficial aspects of RFID tags such as their batch readability, resistance to harsh environmental conditions, informa- tion storage and processing capability, among others. Given the recent explosion of interest in RFID tag incorporation in supply chains, literature in the area has not yet comprehensively identified nor addressed associated challenges and impediments to successful implementations. We purport to fill this gap and to raise awareness by identifying and discussing critical issues such as ownership transfer, privacy/security, computing bottleneck, read error, and cost-benefit issues such as opportunity cost, risk of obsolescence, information sharing, and inter-operability standards. European Journal of Information Systems (2009) 18, 526533. doi:10.1057 ejis.2009.41; published online 10 November 2009 Keywords: RFID; item-level information; information revelation; information sharing Introduction RFID is a disruptive technology that is here to stay at least for the immediate foreseeable future based on currently available technology for relatively cheap and generally reliable tracking and tracing of items throughout supply chains. Several Special Issues of journals including IEEE Systems Journal (1(2) December 2007), IEEE Transactions on Automation Science and Engineering (6(1) 2009), International Journal of Production Economics (112(2) April 2008), Journal of Theoretical and Applied Electronic Commerce Research (3(1) 2008), Production and Operations Management (16(5) SeptemberOctober 2007), European Journal of Information Systems (2009) have been dedicated to Radio-Frequency Identification (RFID)-related issues. Moreover, the number of conferences and workshops that are dedicated specifically to address various facets of RFID research has exploded over the past few years, and these conferences and journal Special Issues certainly attest to RFIDs relevance. Clearly, EJIS, as a prominent journal in the forefront of cutting-edge Information Systems research, has several publications (including Lee & Shim (2007), and the Special Issue that is due to be published in 2009) that address RFID-related issues. During the past few years, several leading retailers including Wal-Mart, Metro Group, and Marks&Spencer have mandated the use of RFID tags for tracking items primarily at the pallet level. Apparently, Wal-Mart claims to be able to achieve $287 million benefit by fixing just 10% of the problems associated with lost sales due to misplaced inventory (Nystedt, 2007). Although not many RFID implementations have resulted in benefits that European Journal of Information Systems (2009) 18, 526533 & 2009 Operational Research Society Ltd. All rights reserved 0960-085X/09 www.palgrave-journals.com/ejis/ are in that order of magnitude, it is worth noting that given the steady decrease in their implementation cost, the benefits they provide, and the need to maintain competitive advantage, RFID tags are here to stay. The number of RFID tags produced each year is a small fraction compared to that of bar codes. For example, Das & Harrop (2008) expect that 2.16 billion tags will be sold in 2008 whereas about 510 trillion bar codes are printed every year. However, it is generally agreed that in the very near future RFID tags would supplant bar codes. While RFID tags are certainly poised to supplant bar codes (e.g., pallet tags in supply chains) in a good number of supply chain applications, their capability and poten- tial applicability spans an even wider range (e.g., sensor tags) vis-a`-vis bar code applications. RFID tags have been known to provide several benefits for supply chain management including improved asset management, improved inventory control, shrinkage reduction, increased product availability and fulfillment rates, reduction in labor cost, decreased operational time, material handling efficiency, improved process through- put, improved customer service including returns & recall management as well as service and warranty authorizations, increased in-transit visibility, confirming regulatory compliance through chain-of-custody records, and overall process improvement. Benefits that are measured as short term or long term can be either tangible or intangible. RFID implementations, like several technological implementations, do also exhibit network effects simply because their value is better realized when several related entities provide and utilize the necessary infrastructure. Their beneficial properties, therefore, are only bound to improve over time as RFID tags gain widespread use. Although a majority of related literature consider the beneficial aspects of RFID implementations in supply chains, we are not aware of any that specifically considers the deterrents in a supply chain management context in-depth. Lee & Ozer (2007) observe that a large gap exists between perception and reality of what is generally assumed to be the benefits engendered by the incorpora- tion of RFID tags in supply chains. When one considers various research issues that have been addressed related to RFID, including those with a focus on supply chain management, a majority of them either ignore or assume away the complexities that are associated with RFID implementations. To our knowl- edge, no existing published work attempts to flesh out the details on the major challenges that one faces in RFID implementations for supply chain management. Our purpose, therefore, is to raise awareness by identify- ing and comprehensively presenting several critical issues that need to be considered and addressed before wide- spread adoption of RFID becomes a reality. Given the initial resistance to any new technology, of which RFID is not immune, critically examining the risks, benefits, and associated challenges sooner than later would facilitate seamless introduction and use of RFID tags in supply chains. To this end, and in line with the Editors view (Paul, 2005, p. 208), we identify and discuss specific issues related to ownership transfer, cost, privacy/security, back-end systems, obsolescence, read error, economic disincentives to share item-level information, and the complex nature of evolving standards that address various facets of RFID tag applications. The following sections consider them one by one, and we conclude the paper with a brief discussion. Ownership transfer issues In a traditional supply chain, ownership transfer occurs through physical means where the seller no longer has (physical) access to the sold item without the buyers knowledge. However, when RFID tags are embedded in the sold item, the previous owner can indefinitely maintain (radio-frequency, RF) item-level access to the sold item due to the characteristics of RFID tags to be read and written through wireless means. While a previous owner may not have physical access to the sold item, the item can be tracked and traced through RF means and this has the potential for security/privacy violations as well as loss in competitive advantage. Security/privacy violations occur when the item is tracked and traced without knowledge of (and explicit or implicit permis- sion from) the current owner. As RF communications may occur without physical proximity, it is relatively easy for the previous owner to surreptitiously verify the location of the tagged object. Knowledge of the presence of specific tagged items can be used to generate informa- tion (e.g., inventory level) that directly affect competitive advantage. Complete (i.e., both physical and RF) owner- ship transfer is therefore critical when competitive advantage, security/privacy issues, etc. are at stake. Researchers have thus far attempted to address this issue of complete ownership transfer in the presence of RFID tags through cryptographic means. When complete ownership transfer occurs, the previous owner should not have any (including RF) access to the RFID tag without permission from and knowledge of the new owner. After extensively reviewing existing literature, Kapoor & Piramuthu (2008) find that a majority of existing cryptographic approaches are vulnerable to attacks by adversaries, which may possibly include previous owners. These adversaries, with appropriate and necessary re- sources, are able to obtain usable information from passively observing or even actively participating in the conversation between an authentic tag and reader. This information can then be used to track, trace, or even impersonate the authentic reader or RFID tag. Such infor- mation can also be used to generate Denial of Service (DoS) attacks, with the ability to cause extensive disrup- tions and loss in the system. Kapoor & Piramuthu (2008) also observe that none of the existing cryptographic approaches accomplish ownership transfer without the presence of a trusted third party (TTP). Although it is feasible to use a neutral third party such as a bank as a TTP to facilitate ownership transfer, the process can get Challenges associated with RFID tag implementations Gaurav Kapoor et al 527 European Journal of Information Systems unwieldy due to the presence of different TTPs between each buyerseller pair in a supply chain and coordinating cryptographic secrets among these TTPs is not a trivial task. Therefore, it is impractical to assume that TTPs will be able to coordinate every transaction in a supply chain. Moreover, even if TTPs are assumed to be present at every ownership transfer instance, the process is rather difficult to accomplish simply because the TTPs need to have some mechanism to share the tags updated secret keys as and when ownership transfer occurs. Given the huge volume of RFID tags, each with its own unique instance-level information and the movement of the tags in and out of the field of the reader over time, it is not feasible to expect every TTP to have access to the next TTP in the ownership transfer sequence. The TTPs cannot communicate with one another unless there is some sort of mechanism through which they can authenticate themselves to one another. As seen in Figure 1, the tag has to (wirelessly) comm- unicate with current and previous owners as well as the TTP to exchange secret key information. Such wireless communication are vulnerable to attacks by adversaries (represented by the shaded triangle in Figure 1). These attacks have the potential to reveal the tags secret key information. And, once the secret key information is revealed, the adversary can track and trace the tag or even impersonate tag to reader or reader to tag. Even if the tags secret key information is not revealed, the tag becomes vulnerable as soon as the adversary recognizes some pattern (i.e., signature) that is specific to this tag that can be used to track and trace. An active adversary can modify communication between tag and reader to cause DoS attacks, which result in permanent loss of communication between tag and reader. Without TTP, the scenario is even worse. No existing ownership transfer protocol guarantees complete owner- ship transfer (Kapoor & Piramuthu, 2008). Existing ownership transfer protocols without a TTP accomplish only ownership sharing, where every previous owner continues to maintain RF access to the tags. This, clearly, is not acceptable in a supply chain context unless com- plete information visibility is agreed upon by every node in the supply chain. Moreover, this is exacerbated even further by the deleterious effects of authentication protocol vulnerabilities that can be taken advantage of by adversaries. Realistically, it is impossible to develop an ownership transfer protocol without a TTP, as any key update information that is broadcast through a wireless medium between (previous and new) owners and RFID tag can be read by anyone in the vicinity. With currently available technology, the best that can be accomplished in the absence of a TTP is ownership sharing whereby the previous owners continue to maintain RF access to the tag. Given the absence of TTP in several prospective RFID applications and the sheer difficulty of cryptographically transferring ownership under these circumstances, it is surprising that this aspect of RFID implementation has received very little attention in related literature. The lack of mention of this aspect is even more surprising given the fact that ownership sharing is generally not what one expects when ownership transfer takes place, as the former could lead to privacy/security violations and can deteriorate competitive advantage for the current owner. Cost issues As in bar code applications, the cost associated with RFID applications in supply chains include the cost to obtain and tag items, tag readers, as well as necessary back-end systems to gather, maintain, and process the data including changes to existing Enterprise Resource Plan- ning (ERP) and other related systems. While it is relatively straightforward to compare the unit cost of RFID tags and bar codes and their respective readers, the rest of the cost structure is not necessarily obvious. This could be due to the rarity of RFID implementations in supply chains (Murphy-Hoye et al., 2005). As more and more imple- mentations become available, the experience gained from these would provide the necessary information to gen- erate informed estimates of costs and associated returns on investment (ROI). It is clear that RFID implementations are more expensive than a comparable bar code implementation. However, the benefits of RFID implementations through reducing labor, out-of-stocks, shrinkage, etc. are expected to far exceed costs (e.g., Das & Harrop, 2008). These benefits may only be realized if the RFID implementa- tions are fully utilized. That is, an upstream firm imple- menting slap-and-slip to comply with mandates from firms downstream in a supply chain may end up with just the costs and none of the benefits unless it strategically and tactically uses readily available information from those RFID tags. For example, such a slap-and-slip implementation just to comply with Wal-Marts mandate is estimated at $500,000 for a top supplier (Maurno, 2005). Other deterrents to adoption include the perceived lack of ROI and the fact that an ROI model for one firm may not be directly transferable for another to leverage. The problem is compounded when payback time is not short, as is certainly the case with most typical RFID investments. Owner 1 Owner 2 Owner 3 Owner n T TP 1 TTP 2 TTP 3 TTP n-1 Tag Tag Ownership Transfer Tag RF Wired Figure 1 RFID tags progress through a supply chain. Challenges associated with RFID tag implementations Gaurav Kapoor et al 528 European Journal of Information Systems It should also be noted that, in an overwhelming majority of published literature in this area, the cost of bar codes is deflated to cover just the cost of printing them. The reality is that unlike RFID tags, which can be embedded in an item (e.g., inside the casing) bar codes have to be printed on the outside of the item where it is exposed to the elements. In addition to the difficulty of reading bar codes in different environments (e.g., direct sunlight), replacing a destroyed bar code can be expen- sive since it includes the cost of identifying the item while it is in process, printing a replacement bar code, and returning the item to its place in the process flow. As human input is necessary in several of these steps, the likelihood of error increases, which can result in immeasurable loss. Therefore, a major impediment of RFID adoption is the vagueness of its cost-benefit analysis, which includes both unknown cost structure and unclear future payoff. In general, the cost associated with RFID adoption usually includes cost associated with infrastructure installment, marginal cost on various tags and receivers, and oppor- tunity cost. While the installment cost and marginal cost are relatively easy to measure, the opportunity cost is usually rather difficult to quantify. Privacy/security issues Privacy issues are major stumbling blocks to RFID implementations. Consumer protection organizations such as CASPIAN (www.nocards.org) are vehemently against the spread of RFID tags in consumer applications, as they are miniscule and can be covertly attached to any object that can in turn be inconspicuously tracked and traced without the knowledge of the objects owner. Moreover, it does not take much effort to associate tags with their owners demographic and other information to violate their privacy and security. There are several examples of firms tagging their items only to face extreme opposition from customers (e.g., www.boycottbenetton. com). The information trail left by RFID raises serious privacy/security concerns. There have been several attempts at alleviating secur- ity/privacy concerns. For example, the Blocker Tag (Juels et al., 2003) helps consumers manage their live RFID tags in a privacy-protecting manner. When present and activated, the Blocker Tag deters readers from obtaining information on RFID tags in the vicinity. However, it also prevents communication between authentic tags and readers. Moskowitz et al. (2007) proposed the Clipped Tag, which enhances consumer privacy by shortening (clipping) the antenna in the RFID tag thereby reducing its read range. Although the Clipped Tag works well, in principle, if the tag is no longer needed, it fails when the tag is expected to operate normally at a later point in time. For example, the customer of an RFID-tagged item may have clipped the tag, but later decides to return the item to the store. The item is no longer visible to the store as much as it used to be before clipping occurred and the utility of RFID tag on this item drops precipitously from the stores perspective. A customer may also want the tag to be operational after the item is placed in a refrigerator at home to keep track of its inventory, etc. but not in transit between the store and home. Disabling tags is clearly not universally appropriate. There is a need to empower customers with some form of opt-in/opt-out choice process without any debilitating effect on the smooth operation of existing systems. Researchers have also developed a large number of security protocols that can be used for secure commu- nication between tag and reader (e.g., Piramuthu, 2008). However, this is a very active and evolving area and while new protocols are being introduced with increasing frequency, vulnerabilities present in extant protocols are continually identified. Although researchers proposing some of these security/privacy protocols go the extra mile in proving that their protocols are extra secure (e.g., Juels & Weis, 2005) only to be shot down soon thereafter by someone who identifies a loophole in the security proof (e.g., Gilbert et al., 2005). Clearly, a security/privacy protocol remains secure only until someone identifies its vulnerabilities. Back-end system bottleneck Existing systems that gather, maintain, and process bar code information can, in principle, be used to gather, maintain, and process information from RFID tags. However, the process may not scale up smoothly, as a majority of existing systems are known to be operating at close to their full capacity and any additional load on them may drastically slow down their operation even in the best case scenario. In the worst case scenario, they could completely shut down the system due to processing overload. RFID tags are notorious for generating huge amounts of data as compared to bar codes. For instance, bar codes have class-level information and the frequency of scanning bar codes is generally quite low. RFID tags, on the other hand, embed in them item-level information in greater detail and they are relatively more frequently scanned in most applications. RFID tags, therefore, generate more information per scan and they are scanned more frequently. This leads to an exponential growth in the volume of data. RFID-tagged systems typically generate about 10100 times the volume of data generated in bar code systems. In large retail and supply chain organizations, this trans- lates to daily data generation in terms of terabytes. It is estimated that when item-level tags are used, Wal-Mart is expected to generate about 7terabytes of data every day. The difference is significant even if polling frequencies of RFID tags and bar codes are the same. Its worse when RFID tags are more frequently scanned and when active tags are used. Mobility, as happens in any supply chain, oftentimes necessitates increased frequency in tag reads. Data collected are of no use if not used properly. Most data mining systems that are used to identify patterns in bar code data are generally not scalable, and they are not developed to process data at such huge Challenges associated with RFID tag implementations Gaurav Kapoor et al 529 European Journal of Information Systems volumes. The systems in place for handling bar code data at present, therefore, need to be upgraded to handle the additional data volumes when RFID tags are introduced. Without appropriate use of data generated by RFID tags, an RFID system could end up being an expensive bar code system. Computing resources such as memory, communication bandwidth, and computer hardware are generally of primary concern when an organization that adopts RFID systems has limited budget to upgrade the supporting back-end systems. However, oftentimes it is the compu- tational time that hinders the applicability and expend- ability of an RFID system when the need for data storage and processing time increases exponentially. Consider the simplistic scenario where the data are stored at n different locations (location here refers to the scenario of a server that may be located in a centralized building or spread over different geographical regions), adding one attribute with space requirement of a units will result in na units of additional storage requirement. If attributes are used in m hierarchical computing pro- cesses, the required computing time is in the order of n m (Figure 2). Moreover, in addition to increased number of attributes, introducing RFID tags in supply chains would likely also result in more frequent polling of these tags. These result in more attributes as well as instances corresponding to each tag in the system. In a classical database analysis context, in addition to additional storage space requirement, this translates to an exponen- tial increase in required processing power. Figure 2 illustrates that without evolutionary computa- tional breakthrough, such as quantum computing, current computing technology is only able to increase its computing power in a linear manner by stacking up processing units. RFID data processing requirements, on the contrary, demand exponentially increasing proces- sing power with an increasing scale of information. The computing bottleneck occurs when scale of information and associated computing demand are large. Risk of obsolescence The technology associated with RFID-related security/ privacy protocols and industry standards have been evolv- ing rapidly over the past decade. It is natural for potential RFID adopters to consider the risk of obsolescence as well as the uncertainties in future payoffs. On the other hand, it is not uncommon for newly introduced applications on existing RFID systems to generate unexpected benefits to the organizations involved (Hendricks, 1992; Tellis & Golder, 1996). Consequently for business practitioners who are risk averse, it is difficult to make decisions when the payoff for RFID projects is extremely unpredictable. For example, the current system may be obsolete in a short period of time or additional unexpected benefits may result from unplanned new advances (e.g., in technology) in the future. While the problem with the risks associated with future payoff is a common issue to any new technology investment scenario, there are some unique character- istics associated with RFID adoption. Firstly, an RFID project usually requires continual investment over a considerably long time span and failure could happen at any time. Secondly, the uncertainty of future payoff is very volatile compared to most other new technologies. The causes of this large variance include (1) potential obsolescence; (2) possible unplanned advances (in tech- nology, among others); and (3) uncertain adoption com- patibility from supply chain downstream and upstream. Although the actual payoff may, also, be better than the expectation value, the risk associated with RFID adop- tion can, nevertheless, not be ignored. A Real Options approach may be appropriate in this context, given the risks associated with RFID investments (Figure 3). RFID Data Processing Time Requirements Computing Power Bottleneck Figure 2 Comparison of the speed of growth of RFID data processing requirement and (computing) processing power. I 1, Initial Investment P 11 P 12 t 0 G 12 , Cease Investment G 24 , Cease Investment P 21 P 22 P 23 P 24 P 31 P 32 P 33 P 34 P 35 P 36 P 37 P 38 G 32 G 34 G 35 G 36 G 37 G 38 G 33 G 21 , 3rd Phase Investement G 11 , 2nd Phase Investement G 22 , Cease Investment G 23 , 2nd Phase Investement G 31 t 1 t 2 t 3 Figure 3 RFID investment decisions across time. Challenges associated with RFID tag implementations Gaurav Kapoor et al 530 European Journal of Information Systems Here, the implementer has the option to bail out or proceed further with investments as situations dictate over time. Read error A majority of the literature on RFID assume that once RFID tags are affixed on items, they can be tracked and traced without any problem. Although it is generally assumed that data read from RFID tags are highly accurate, variations in accuracy can and do occur due to several reasons. Jeffery et al. (2006) claim that over 30% of RFID tag reads are routinely dropped. Although the extent of dropped reads varies across domains and are context-specific, figures in the 6099% range are not uncommon. These read rate errors could occur due to several reasons including the presence of RF-impenetrable (e.g., metallic) objects between tag and reader, inconsistencies across different readers, chip and antenna defects, noise, reader and/or tag collisions, among others. While some of these can be prevented or their deleterious effects can be alleviated, it is difficult to guarantee a 100% read rate under any circumstance due to the dynamic nature of the process of reading RFID tags as well as their ambient conditions. A tag read could result in several different scenarios true positive, true negative, false positive, or false negative. True positive (or, true negative) refers to the case where the reader confirms the tag is present (or absent) when the tag is actually present (or absent). False positive refers to the case when the reader confirms that the tag is present when indeed it is not. False negative refers to the case when the reader confirms that the tag is absent when indeed it is present. In a supply chain management context, false positives and false negatives generate their own distinctive dynamics. For example, too many false positive reads could lead to stock-out situations resulting in related loss in revenue and customer goodwill. Too many false negatives, on the other hand, for example, could result in excessive inventory holding cost and related loss due to spoilage, obsolescence, etc. Several researchers (e.g., Tu et al., 2009) have attempted to address this issue with varying degrees of success. Clearly, RFID tag read error cannot be completely eliminated. However, to benefit from detailed item-level information provided by RFID tags (e.g., Zhou, 2009), the read rate error needs to be reduced to a minimum. Economic disincentives to share item-level information Conflict of interests between sellers and buyers to share RFID-generated information is another major impedi- ment to RFID adoption in supply chain management. For example, Wal-Mart requires its vendors to be equipped with up-to-date RFID infrastructure according to Wal-Marts compatibility standards. While Wal-Mart will definitely benefit from shared information from up- stream, vendors are reluctant to share such information, which traditionally is kept as private, even without considering the extra cost associated with the revelation of such information. As it is not uncommon for a supplier to supply items to several buyers, sharing of RFID-generated information (including, for example, inventory information) could lead to serious deteriorations in competitive advantage. Although the nodes upstream in a supply chain do benefit from reduction in variance (bullwhip effect) across the supply chain, the nodes further downstream benefit the most. The irony is that the nodes further up in the supply chain are the ones who incur a majority of the costs associated with RFID tagging, etc. while they generally benefit the least from such implementations. Given this scenario, the nodes upstream have only minimal incentives, if any, to share item-level informa- tion with nodes downstream. That is, only the categorical information is transferred along with the sold item. The seller generally provides only aggregate information on the sold items such as minimum quality standard (MQS), whereas the buyer bears the cost associated with obtaining the minimum required sample and verifying that these MQS are met. An advantage when RFID tags are embedded in these items is that the specifications are transparent, and therefore preclude the need for sampling. Economic disincentives to share RFID information, nevertheless, has become one of the major concerns for companies in a supply chain to adopt RFID technology when the benefits are perceived to be only marginal. Evolving standards As with any widely adopted technology, regulatory stan- dards addressing various facets of RFID applications are necessary to ensure minimum level of product character- istics including quality, inter-operability, reliability, and safety. While there are different standards and guideline developers at the international (e.g., ISO), national (e.g., ANSI, BSI), pan-industry (e.g., EPCglobal, AIM Global), and industry-group (e.g., AIAG, Air Transport Association) levels, the effects of globalization necessitates that these standards are compatible for RFID tag applications. Given that there are several different types of RFID tags and readers that operate across national boundaries and applications, universal standards that address the entire application spectrum are necessary to improve the inter- operability of these standards. RFID standards generally address issues related to tag and reader frequency, air interface protocols, data structure, data content and format, tag and reader types, tag and reader conformance, code assignment, and data communications. The Global RFID Interoperability Forum for Standards, funded by the European Commission to improve colla- boration and maximize global interoperability of RFID standards, has identified more than 125 RFID standards. However, there is a lack of consistency among these (e.g., EPCglobal and AIM Global have their own emblems for indicating that a particular product or its packaging Challenges associated with RFID tag implementations Gaurav Kapoor et al 531 European Journal of Information Systems contains an RFID tag). Supply chain participants and standards developers need to be aware of the differences and overlaps among existing standards and be able to create new standards when necessary while ensuring that these interoperate with existing standards. For example, several ISO/IEC standards exist for RFID applications including ISO/IEC 18000, ISO/IEC 15961-3, ISO/IEC 17363-7, and ISO/IEC 18185. The ISO/IEC18000 series was developed to determine the use of the same air interface protocols for RFID item management. Specifically, ISO/IEC 18000-3 defines the air interface at 13.56MHz to be used for a single read from a short distance that is ideal for personal identification applica- tions, the ISO/IEC 18000-6 is for batch-reading from medium-range read distance that is ideal for high-speed object identification in a general supply chain context. The DASH7 Alliance supports broader use of ISO18000-7 standard to track a wide range of objects including vehicles, shipping containers, pharmaceutical products, hazardous materials, perishable goods, and manufactur- ing and operational equipment so the tagged object can move through the supply chain without any disruption. The Health Industry Business Communications Council recently approved a set of standards for RFID tags for labeling and tracking medical products (ANSI/HIBC 4.0) that include guidelines (e.g., the use of 13.56MHz high- frequency coding) for tagging healthcare products to prevent radio frequency interference with medical devices. In addition to the technology standard, the regulatory aspects differ among countries. These differences need to be addressed for widespread RFID adoption in supply chains and to alleviate related barriers to global trade. Moreover, from a supply chain perspective, most stan- dards address the container level or pallet level handing of materials and there is a need for standards that address tagging at the item level. This is an evolving process, and new standards are continually being proposed, evaluated, and approved. Nevertheless, the lack of and inconsisten- cies among standards only slow down the rate of adop- tion of RFID tags in general supply chain applications. Conclusion We considered, and hopefully raised awareness on, several critical issues that can possibly hinder smooth RFID implementations in supply chains. Research to address these issues is sorely needed and raising aware- ness is a necessary early step in this process. Although it is generally believed that RFID tags will be incorporated extensively in supply chains, it is only a matter of time before this becomes a reality. However, to facilitate smooth transition from bar code implementations to RFID implementations, the issues that we considered need to be addressed. While some of these can be completely addressed or their deleterious effects can be alleviated through appropriate means, others (e.g., own- ership transfer without a TTP) are rather challenging to accomplish with current technology. We conclude by summarizing the many issues facing the RFID industry into two main categories: the technical issue and the economic issue. Although emerging technologies have had continual impact and innovative applications in RFID have been developed over the years, we cannot avoid the fact that inherent technical con- straints have hindered the development and adoption of RFID devices in many industries. Among the technical issues, we considered four of the most significant ones: (1) ownership transfer issue; (2) privacy/security issue; (3) computing bottleneck issue; and (4) read error issue. So far, both academic and industry researchers in this area have devoted a significant amount of their resources to address some of these issues with promising results. We anticipate a steady increase in technical advances as the area gains more prominence in the very near future. Economic issues also have enormous influence on RFID adoption in supply chain management. Among them, we considered four prominent issues including issues related to (1) cost-benefit considerations; (2) opportunity cost considerations and risk of obsolescence; (3) information sharing among nodes in supply chains; and (4) the need for standards that facilitate inter-operability and global trade. Just as the Internet has changed the way people live, the disruptive technology that is RFID will change the way business operates in a revolutionary manner. New business phenomenon, innovative economic strate- gies, and improved business process controls that have emerged over the past several years have been gaining attention from both business sectors and academic litera- ture alike. We believe that by addressing the economic and technical challenges facing this industry, we will be in a better position to help the industry better understand this technology and its untapped potential in supply chain applications. Acknowledgements The authors thank the anonymous reviewers, Associate Editor, and Editor for their helpful suggestions. About the authors Gaurav Kapoor received his Ph.D. in Information Systems from the University of Florida. His research interests include RFID systems. Wei Zhou received his Ph.D. in Information Systems from the University of Florida. He is Assistant Professor of Information Systems and Technologies and a member of the RFID European Lab at ESCP Europe. His research interests include RFID-enabled item-level information visibility, Internet advertising, and knowledge-based learning systems. Challenges associated with RFID tag implementations Gaurav Kapoor et al 532 European Journal of Information Systems Selwyn Piramuthu is Professor of Information Systems at the University of Florida. He is a member of the RFID European Lab at ESCP Europe. His research interests include RFID systems, pattern recognition and its application in supply chain management, computer- aided manufacturing, and financial credit-risk analysis. References DAS R and HARROP P (2008) RFID Forecasts, Players & Opportunities 20082018. IDTechEx, http://www.idtechex.com. HENDRICKS K (1992) Reputation in the adoption of a new technology. International Journal of Industrial Organization 10, 663677. GILBERT H, ROBSHAW M and SIBERT H (2005) An active attack against HB+ a provably secure lightweight protocol. Institution of Engineering and Technology Electronics Letter 41(21), 11691170. JEFFERY SR, GAROFALAKIS NM and FRANKLIN MJ (2006) Adaptive cleaning for RFID data streams. Proceedings of the 32nd VLDB Conference pp 163174, ACM, Seoul, Korea. JUELS A, RIVEST RL and SZYDLO M (2003) The blocker tag: selective blocking of RFID tags for consumer privacy. In Proceedings of the 8th ACM Conference on Computer and Communications Security (ATLURI V, Ed), pp 103111, ACM Press, Washington, DC. JUELS A and WEIS SA (2005) Authenticating pervasive devices with human protocols. In Advances in CryptologyCRYPTO05 (SHOUP V, Ed) Vol. 3126, pp 293308 Lecture Notes in Computer Science, Springer-Verlag, Berlin, Germany. KAPOOR G and PIRAMUTHU S (2008) Protocols for objects with multiple RFID tags. The Sixteenth International Conference on Advanced Comput- ing and Communication (ADCOM2008). pp 208213, IEEE Press, Chennai. LEE C-P and SHIM JP (2007) An exploratory study of radio frequency identification (RFID) adoption in the healthcare industry. European Journal of Information Systems 16, 712724. LEE H and OZER O (2007) Unlocking the value of RFID. Production and Operations Management 16(1), 4064. MAURNO DA (2005) Going for (not so) broke: the true cost of RFID. Inbound Logistics 25(7), [WWW document] http://www.inboundlogistics .com/articles/features/0705_feature05.shtml (accessed 25 August 2009). MOSKOWITZ P, MORRIS S and LAURIS A (2007) A privacy-enhancing radio frequency identification tag: implementation of the clipped tag. Proceedings of the Fifth Annual IEEE International Conference on Pervasive Computing and Communications Workshops (PerCom07) pp 348351, IEEE Computer Society, White Plains, NY. MURPHY-HOYE M, LEE H and RICE JB (2005) A real-world look at RFID. Supply Chain Management Review 9(5), 1824. NYSTEDT D (2007) Wal-Mart eyes $287 million benefit from RFID. IDG News Service, 12 October. PAUL RJ (2005) Editors view: an opportunity for editors of IS journals to relate their experiences and offer advice. The editorial view of Ray J Paul. European Journal of Information Systems 14(3), 207212. PIRAMUTHU S (2008) Lightweight cryptographic authentication in passive RFID-tagged systems. IEEE Transactions on Systems, Man, and Cybernetics Part C 38(3), 360376. TELLIS GJ and GOLDER PN (1996) First to market, first to fail? Real causes of enduring market leadership. Sloan Management Review 37, 6575. TU Y-J, ZHOU W and PIRAMUTHU S (2009) Identifying RFID-embedded objects in pervasive healthcare applications. Decision Support Systems 46(2), 586593. ZHOU W (2009) RFID and item-level information visibility. European Journal of Operational Research 198(1), 252258. Challenges associated with RFID tag implementations Gaurav Kapoor et al 533 European Journal of Information Systems