| |
|
|
RFID
Deployment
Considerations for Accountants
By P. Paul Lin and Kevin F. Brown
AUGUST 2008
- Radio frequency identification (RFID) is transforming the way
organizations track inventory and plant assets, and will soon allow
companies to automate business processes and asset management. Many
large retailers are focusing on deploying RFID in their supply chain
management, and most are probably only a few years away from item-level
RFID deployment in their stores. As RFID is increasingly integrated
into enterprise information systems, accounting professionals must
be well-acquainted with the technology, specifically deployment
issues including technology standards and RFID middleware.
Current
Barcode Systems
The majority
of today’s business information systems are designed to
handle barcodes. However, barcode systems are not designed to
track individual inventory items, as they are typically assigned
to product categories. For example, most merchandise sold in the
United States carries a 12-digit universal product code (UPC)
approved by the Uniform Code Council (UCC). The UPC contains the
product type (the first digit), the manufacturer’s code
(the next five digits), the product code (the five digits following
the manufacturer’s code), and a check digit (the last digit).
The publishing industry uses the 13-digit European Article Numbering
(EAN) system for encoding ISBN numbers on books. While these popular
systems can identify product categories, they do not have the
capability to uniquely identify individual items.
When merchandise
is scanned at the checkout counter in a barcode-based retail environment,
the barcode data are transmitted to the store’s computer,
which instantly locates the matching record in the master file.
The matching record contains an array of important information,
including product description, sales price, preferred vendor,
quantity on hand, and reorder point. Upon the sale of merchandise,
the barcode system will subtract one unit from the quantity on
hand and update the master file. Conceptually, the process is
similar to the lookup functions used by accountants in Microsoft
Excel.
RFID
Standards
In contrast
to category-level barcodes, RFID systems can treat every inventory
item individually. To use this capability, however, the technology
requires industry-wide standards. The EPCglobal consortium, a
nonprofit organization, is a collaboration between EAN and UCC
to promote the adoption and standardization of electronic product
code (EPC) technology. Exhibit
1 shows EPCglobal’s outline of a network of EPC-enabled
data services used by companies to enable near-real-time information
retrieval for items in their supply chains (Himanshu Bhatt and
Bill Glover, RFID Essentials, proquest.safaribooksonline.com/0596009445/rfid-CHP-8-SECT-2).
Such an RFID system is composed of up to three subsystems [Guidelines
for Securing Radio Frequency Identification (RFID) Systems,
National Institute of Standards and Technology, April 2007]:
- The
RF subsystem, consisting of EPC tags and readers,
performs identification and related transactions using wireless
communication.
- The
enterprise subsystem is composed of network infrastructure,
middleware, databases, applications systems, and web servers.
The enterprise subsystem stores, processes, and analyzes data
acquired from RF subsystem transactions to make the data useful
to support business processes.
- The
inter-enterprise subsystem is composed of Internet,
object naming service (ONS), and discovery service. The inter-enterprise
subsystem connects enterprise subsystems when information needs
to be shared across enterprise boundaries (e.g., between customers
and suppliers).
Every RFID
system contains an RF subsystem for reading or writing tags and
an enterprise subsystem to assimilate the information into business
processes. An inter-enterprise subsystem is needed to support
the data exchange required in supply chain management systems.
EPCglobal
has more than 1,000 members worldwide and its board of governors
includes Walmart, Procter & Gamble, Johnson & Johnson,
Cisco, Hewlett Packard, and the U.S. Department of Defense. EPCglobal
functions as the standards setter for the RFID industry. Exhibit
2 shows the standards already ratified by EPCglobal to date.
Note that these standards are subject to continual updating by
EPCglobal.
The steady
decline in the price of tags has accelerated RFID implementation.
These tags are categorized into three classes:
- Class
0. Data are encoded when tags are manufactured (i.e., read-only).
- Class
1 Gen 1. Data are encoded in tags once by users.
- Class
1 Gen 2. This latest type of tag can operate globally in the
United States, Europe, and Asia in the 860- to 960-MHz UHF band.
A Gen 2 tag has a memory size of 96 bits or larger.
With a minimum
of 96 bits, the memory of Gen 2 tags can accommodate the item-level
tagging needs of any company. The 96 bits are divided into four
segments: header (8 bits), EPC manager (28 bits), object class
(24 bits), and serial number (36 bits). Given the binary representation
for every bit, a total of 28 bits for EPC manager can represent
up to 268 million companies, and the object class can represent
up to 16 million product groups for any given company. Moreover,
the 36 bits reserved for serial numbers can represent up to 68
billion different identities in the same product group. The data
representation capability of a Gen 2 tag is sufficient to provide
unique identifiers for all items produced worldwide. This enormous
data identification capacity, however, requires vast amounts of
storage to accommodate the volumes of data caused by RFID deployment.
RFID
Middleware
Business
applications of RFID have only just emerged in recent years, so
most enterprise information systems are not ready to directly
process the data collected by RFID readers. Consequently, the
deployment of RFID requires middleware, which provides the interface
between RFID readers and a company’s existing application
systems. The three primary reasons for using RFID middleware are:
1) to provide connectivity with readers while encapsulating the
applications from the device interface and interconnections; 2)
to lower the volume of information that applications need to process
by filtering and grouping raw RFID observations coming from readers;
and 3) to provide an application-level interface for managing
readers and querying RFID observations. RFID middleware can apply
filters and logic to tag data collected by readers so that business
events can be properly processed by other application systems
(Himanshu Bhatt and Bill Glover, RFID Essentials, proquest.safaribooksonline.com/0596009445/rfid-CHP
-2-SECT-5#X2ludGVybmFsX1RvYz
94bWxpZD0wNTk2MDA5NDQ1L3JmaWQtQ0hQLTc=).
Middleware
providers can be grouped into three categories: large IT conglomerates
(e.g., HP’s OpenView and IBM’s WebSphere), software
vendors (e.g., Microsoft’s BizTalk Server, SAP’s NetWeaver,
and Oracle’s Fusion Middleware), and pure RFID vendors (e.g.,
OATSystems and Intermec). Companies considering RFID deployment
must weigh the pros and cons of the various vendors before selecting
one to provide RFID solutions. Because the item-level RFID deployment
will result in an explosion in the volume of data processed and
stored by an information system, IT conglomerates offer middleware
with the intent to complement and promote their other products
or consulting services. When
choosing middleware from an IT conglomerate, a company might simply
try to integrate RFID data into its existing application systems.
The company, however, risks missing the greater payoff of RFID
technology: the opportunity to enhance its business processes.
Software vendors often provide proprietary middleware in order
to retain or expand their market share. Consequently, when working
with a software vendor, a company will likely align its RFID deployment
with the vendor’s enterprise resource planning (ERP) software
product lines. Pure RFID vendors often work closely with other
hardware and software providers on various projects; therefore,
they may be biased toward a particular vendor. When making a middleware
choice for deployment, companies should select vendors that are
familiar with their business operations and enterprise information
systems.
The rapid
evolution of RFID products has resulted in a wide variety of RFID
tags and readers of various types and signal frequencies. EPCglobal’s
standards-setting efforts are having a significant impact on RFID
middleware development. Compliance with EPCglobal standards ensures
the interoperability between middleware and RFID products. Furthermore,
the recently ratified Application Level Events (ALE) standard
ensures a standardized interface to access filtered tag data.
Several vendors have already introduced their ALE-compliant middleware
in the market, without which application developers would have
to write additional programs for communicating with devices, filtering
data, and assimilating data to application systems. The ALE standard
provides simple yet flexible mechanisms to filter and group RFID
data. This filtering and grouping capability provides a means
to identify and focus on relevant business events (Himanshu Bhatt
and Bill Glover, RFID Essentials).
According
to one RFID deployment expert, many early RFID adopters later
regretted their haste in selecting the middleware for their RFID
system (Mary Catherine O’Connor, “How to Choose RFID
Middleware,” RFID Journal, August 13, 2007, www.rfidjournal.com/article/articleprint/3511/-1/1).
While their initial middleware selection may have suited their
earlier needs, these companies are discovering, as they attempt
to more fully integrate their RFID system with their business
processes, that their systems lack flexibility. As a result, many
companies are now ripping out their old RFID platforms in favor
of newer, more expensive platforms. The above expert suggests
that companies may be wise to start their RFID deployment on a
small scale, but adopt more flexible middleware from the outset,
even if it is pricier than less-flexible alternatives. More careful
planning, including a better analysis of future RFID integration,
may result in substantial savings.
Additional
Caveats
Businesses
are moving toward a world of item-level identification in an RFID
environment, where the EPCglobal standards will drive tagging
mechanisms, transmission protocols, and network formation (“Discovery
Services—Enabling RFID Traceability in EPCglobal Networks,”
IBM; www.almaden.ibm.com/cs/projects/iis/hdb/Publications/
papers/comad2006.pdf). IDTechEx, an RFID research firm, forecasted
that retail eventually will be the biggest market for RFID; it
also predicted that the real opportunity lies with item-level
tagging, as opposed to the current pallet-level tagging some retailers
have already adopted (“Benefits of Case/Pallet Tagging Rapidly
Flow to Retailers,” IDTechEx; www.glscs.com/news/focus.htm).
Nonetheless, opportunities will abound in almost all business
environments. Decisions regarding when and how to adopt RFID technology
will generally be determined by the tradeoffs among several factors,
including cost, transmission rate, transmission range, and potential
environmental interference. No single RFID technology has proved
to be superior in all applications. Incidentally, the diversity
of IT infrastructures has become a hurdle to RFID deployment because
companies are having trouble deciding which technology offers
the best long-term solution.
The volume
of event data collected in the RFID environment is enormous. In
order to reap the benefits of RFID technology, a huge volume of
data needs to be filtered, processed, and stored in a network
environment. Companies must evaluate the impact of RFID deployment
on data storage, data communication, and information sharing (within
the firm and with business partners) on system infrastructure.
In a barcode environment, a clerk is always near a scanner to
ensure the machine works properly and that barcodes are being
read. Human involvement with the data collection process in an
RFID environment, however, will be minimal or nonexistent. It
is therefore necessary to implement automated means to monitor
and manage RFID readers. The mission-critical RFID network must
be available at all times and continually assessed for potential
system threats. Failure to evaluate an RFID system’s vulnerabilities
and lack of business continuity planning may result in a serious
disruption of operations
Accounting
Applications
The RFID
data filtering and grouping capabilities allow businesses to select
transactions associated with particular events, such as sales,
sales personnel incentives, introductions of new products, and
changes in existing products, for analysis. RFID technology enables
accountants and auditors to tie events to transactions and evaluate
cause-and-effect relationships. While RFID data capabilities are
useful to management, external auditors can use the system as
well, although the massive amounts of data can be burdensome if
they are not effectively filtered.
The item-level
identification offered by RFID can help improve numerous business
processes. Accountants and managers should review their processes
and ask themselves where item-level identification would benefit
decision-making processes, manufacturing processes, and overall
company profitability.
The use of
any IT tool will impact a company’s internal control systems,
as well as the external audit of those systems. The Committee
of Sponsoring Organizations (COSO) defines the objectives of the
internal control process as effectiveness and efficiency of operations,
reliability of financial reporting, and compliance with applicable
laws and regulations (www.coso.org).
Using RFID technology to track inventory and plant assets should
raise concerns about safeguarding resources, under the first objective,
and financial reporting for inventory and fixed-asset data, under
the second.
RFID technology
should be implemented with the goal of increasing the efficiency
and effectiveness of internal control procedures and audits, in
addition to reducing the cost of serving customers while increasing
inventory and plant-asset data collection. A complex IT environment
increases the risks of inaccurate data processing, unauthorized
access to data, and potential loss of data, as well as the importance
of control systems to address those risks. A discussion of IT
internal control procedures is beyond the scope of this article,
and readers can refer to Public Company Accounting Oversight Board
(PCAOB) Auditing Standard 5, as well as the “Preliminary
Staff Views” guidance for smaller companies (www.pcaob.com).
Organization
Implementation
When deploying
RFID technology it is important to assess both the business needs
and system requirements, and to evaluate how these needs will
evolve over time. Instead of treating RFID as a mere data-
collection device, companies should take advantage of RFID’s
real-time information to improve business processes. One formidable
obstacle for RFID deployment is the shortage of RFID expertise.
A report by the Aberdeen Group revealed that more than half of
the companies surveyed suffered from a critical internal shortage
of qualified RFID staff (Evan Schuman, “Major RFID Hurdles
Ahead,” eWeek, July 20, 2006, www.eweek.com/c/a/Retail/Report-Major-RFID-Hurdles-Ahead-[1]/).
A more
recent survey by the Computing Technology Industry Association
found that 84% of technology resellers, solution providers, systems
integrators, and consultants will or might offer RFID products
and solutions in the next three years (“Implementing RFID
Solutions Survey,” TechNews, June 24, 2007, www.technologynewsdaily.com/node/7233).
Companies should hire consultants if they do not have qualified
personnel for RFID deployment in-house. The same survey also indicates
that, despite the slower-than-expected adoption rate, IT firms
remain bullish on the future of the RFID market.
P.
Paul Lin, PhD, is an associate professor of accountancy,
and Kevin F. Brown, PhD, CPA, is an assistant professor
of accountancy, both at Wright State University, Dayton, Ohio.
|
|