Metrics forecast for post implementation of passive RFID technology.
In 2003, the Strategic Distribution (SD) end-to-end distribution process was defined to improve the visibility of the military door-to-door support to the warfighter. A telling remark in that document was that: "The defense distribution business requires more change in the next ten years than it has seen in the last thirty (years)" .
The meaning of this statement still rings true today. However, this ten-year time horizon has shifted focus due to unexpected military events after September 11, 2001, and the realities of ongoing passive radio frequency identification (RFID) pilot tests. The results are an increased focus on the military business process changes, especially in the use of radio frequency tracking and tracing technology. The goal of SD in early 2003 was "to build and provide these improvements via enhanced logistics processes, technology, and better business practices" .
While lean operational improvements such as TQM or Balanced Scorecard continue to contribute to the total asset visibility for the military supply chain system, the focus on the warfighter confidence--itself a metric--seems to remain elusive. Barcode technology has been proven as a tool for total asset visibility since the 1980s. However, the use of passive RFID tags applied to pallets, boxes, containers, and ultimately individual items for the warfighter, still only holds a promise of meeting the confidence metric. Its performance metric has yet to be proven.
How do the assumptions of the SD concept or total asset visibility stand up to the new technology insertion of using RFID? What will be the new metrics, beyond merely reporting dollars per year savings or average number of days in transit?
The emergence of the computer industry in the US and Europe, since the 1950s, demanded new tools and created new metrics for decision making. These computers generated new quantities and assemblages of data, visualizing new patterns of purpose behind those volumes of data. The computer generated new information about manufacturing, supply chains, and operational processes. The insertion of Artificial Intelligence (AI) expert system software, during the 1980s to 1990s, for Army and Air Force logistics systems, started pulling decision-making tools from the custody of the analyst's cubicles into the military's senior executive service offices .
Early expert systems such as PERKS, created by the US Army Logistics Center, were built to "keep decisions consistent" by using computers to govern the personnel requirements for quartermaster and supply services. While these early AI systems, like PERKS, creating a smart AR 350-2, seem trivial today, in 1985, it exploded as a new way of keeping a consistent tracking and tracing of decisions for the generals and senior officials, which lead the way for today's RFID systems keeping track of the supplies and decisions .
The RFID computer chip is the size of a small piece of pepper, which can be imprinted into a metallic ink logo that spells Top Secret, or to a metallic platform, creates an antenna-computer chip assembly shipping label. This small computer chip is like that ubiquitous paperclip collecting bits of valuable information, shown in Exhibit 1 .
The ubiquitous barcode is an automatic identification (or Auto-ID) technology, just like its newest cousin, the RFID code. RFID is similar to the barcode technology, but uses radio waves to capture data from tags rather than optically scanning the bar codes on a label. RFID does not require the tag or label to be seen to read its stored data. That's one of the many differences of an RFID system .
RFID is a disruptive technology. The use of RFID disrupts the old way of operations in logistics and supply chains and business in general (using barcodes).
This disruptive wave of RFID technology is being pushed onto the marketplace by Wal-Mart and the Department of Defense. Both mandated the passive RFID tag be placed on pallets of goods beginning January 2005.
Exhibit 2 shows an old, all too familiar visual cliche of what concerns the military when logistics visibility goes wrong .
ASSUMPTION-BASED PLANNING WAR GAME
RAND developed a planning tool in the 1990s called assumption-based planning (ABP) . This was devised to analyze the basic assumptions for the US going to war with Iraq. This ABP concept would challenge the "most-likely future world extrapolated from current trends" . ABP challenges the assumptions for large programs or decisions, such as mandating the implementation of using passive RFID tags for all military supply pallets and containers.
ABP has five steps: (1) identify the key assumptions; (2) identify the vulnerabilities to these assumptions; (3) define signpost events that signal some possible change to those vulnerabilities; (4) define some shaping action, which is a decision to avert the failure of an assumption that is seen as vulnerable; and (5) define some hedging action, another decision, that prepares the organization for the failure of one of these major assumptions .
The DOD literature regarding total asset visibility actions and the use of RFID technology supplies several key assumptions that seem important in defining future RFID metrics to use in decision making following any total or near total adoption of RFID technology [5,6]. Some of these assumptions are :
* The DOD business base and readiness are diminishing.
* DOD customers (warfighters) find alternative sources and methods of supply to reduce customer (warfighter) wait time.
* Resistance to the use of RFID technology can be overcome by publicizing previous pilot test successes at all levels of the military.
* DOD has demonstrated a reliable distribution chain (with RFID) that can scale upward to the entire DLA or DOD supply chain system.
Challenging these assumptions, and finding others, has to be a priority for the government, if it wants a successful implementation strategy for the passive RFID system.
There are many traditional metrics to track and trace military supply items. Some of the more visible traditional metrics are pallet movement time, the number of pallets (measured in days), pallet movement and receipt segments times (measured in days), average weight of the cargo on these pallets, average cost per pallet for movement, destination of these pallets and cargo, and pallet identification number.
RAND has described the metrics used to track traditional order and ship performance . RAND analysts examined the metrics of average performance and variability in the performance . As in several military reports concerning these performance metrics, the data are suspect in their accuracy. Many times these traditional data do not account for backorders, which could be as much as 15 percent of all requisitions and orders filled. Depending on what time horizon and whether these data are collected during peace time or in time of war, the amount of error could be 700,000 orders missing from analysis for a year, as a minimum .
As more accurate data are collected, possibly with RFID technology, changing these percentages to represent 99 percent to 100 percent as an ultimate goal would be used; or, it will be at least proposed. The ability to capture more data accurately and authoritatively would be a metric design decision to be made during RFID implementation. RFID should allow a better visibility of backorders, which would provide a basis for such an RFID-based metrical decision.
TOMORROW'S METRICS WITH RFID
It is difficult to interpret the impact of this RFID technology on the military decision making process . When we examine any new computer change in the military or civilian application, there is a tendency for the decision makers and the analysts to describe the new features in terms of the current technology being replaced.
The computer has been the hardest technology to predict from all the technologies over the last 100 years that impacted our society or business models. Since WWII, most predictions about the value, the worth, the utility, and the return on investment of new computer technology has been hopelessly flawed, wildly inaccurate .
Two valuable, current RFID metrics for measuring the performance of tracking combat and combat support items are customer wait time (CWT) and requisition wait time (RWT) . The assumption we made was that the military decision makers will continue to require these two metrics until the full implementation of RFID technology. At that point in time, these two metrics will need examining to see if they still fit the needs of military decision makers.
Current performance metrics are time definite measures with quantity defined in units of days, not hours. There are pre-implementation and post-implementation metrics that need to be constantly analyzed.
CWT OR RWT?
CWT measures supply chain performance from the unit perspective: the time it takes to satisfy a request for an item needed directly by the customer. RWT evaluates how well the logistics chain for spare items serves the SSA itself: how much time is required to satisfy an SSA requisition? These requisitions include both those submitted to replenish the SSA's own inventories and those submitted as "special orders" for spare items that are needed by the unit but that the SSA does not stock or have available to issue. Exhibit 3 shows a sample of the variability in the flow of RWT quantity of orders over six months, and Exhibit 4 shows the average processing time in days per RWT segment over six months.
POTENTIAL DATA QUALITY ISSUES
When conducting MILSTRIP data analysis and RWT calculation, we found several exception data points that contribute to faulty analysis. There are indications of potential data quality issues that suggested further clarification is needed from DOD, as well as additional observation and research to be conducted. For our analysis, we found several significant (more than 10 percent) data quality issues that should be addressed as we--DOD--embark on yet another logistics metric research (see Exhibit 5). Four of these issues are:
* Data Issue 1 -- Requisition processing (established) dates were earlier than requisition document dates.
* Data Issue 2 -- MRO processing dates were earlier than their requisition processing (established) dates.
* Data Issue 3 -- The shipment release dates were earlier than their MRO processing dates.
* Data Issue 4 -- The materiel receipt posting dates were earlier than their shipment releasing dates.
The data accuracy problems are real today, even with our new technology and business procedures. Total asset visibility can be improved with passive RFID technology. However, a cost to the military for better data accuracy will be what to do with all that accurate data? What new decisions will generals make when they can see the battlefield logistics flow to the smallest detail, all the time?
RFID METRICS FORECAST
We are in an information-age society and warfare, both defining a new battlespace for logistics . New metrics will emerge for the changing battlespace uses of RFID. We can propose possible new metrics for DOD decision makers, which also should address the above data quality issues.
As RFID has the possibility to capture more data in a timely manner and to do so with potentially few personnel, it is proposed that the following new metrics may need to become a factor:
* Read Rates -- The rate at which the RFID reader records the data from individual RFID tags, and the metrical units.
* Verified Data -- RFID DOD standards or specifications lead toward 100 percent accuracy in reading of the RFID tags on products or items. There may be a need to implement a metric system based on the principles of Six Sigma .
* Relative Asset Visibility -- A value of 1 would indicate constant visibility. A value of 0 would be other visible tracking methods, such as tracking and recording only at a port of entry.
* Hold Time Change -- We expect a change in hold time; possibly a decrease or shorter time.
These and other RFID-focused metrics will emerge from use of this new application of the current and older tracking and tracing policies and technology. The last forty years of military decision makers have shown a demand for new metrics after the new technology is introduced into the military doctrine, policies, and standards .
Without our continued questioning of metrics and assumptions for them, the DOD decision makers' expectations for using RFID and the expected return on investment or benefits may lead to new problems instead of new opportunities for improving total asset visibility.
The challenges to RFID metrics for military, transportation, logistics, and supply chain needs are similar to those of barcodes introduced decades ago.
Todays RFID implementations in the military are demanding a proven measured response. The adoption of RFID as part of a systems thinking paradigm is forcing us to see a set of new metrics to add to the traditional metrics used by the military since WWII.
The transportation components for current order and ship supply items for the DOD are measured in months and days, not hours, minutes, or seconds. Passive radio frequency identification tags could change that model. However, the broad use of RFID is still not being challenged within the DOD logistics systems.
Preliminary analysis of data provided indicates that there are some exceptions to average delivery times measured in days in cargo movement between supply chain nodes. However, this analysis only shows what type of data are available from the military currently. New data capture methods, and data streams will be needed based on RFID tracking to offer a glimpse into the next level of new RFID metrics and metrical units described.
Specific MILSTRIP data analysis and RWT calculation found that several exceptional cases needed to be explored further. The data analysis indicates potential data quality issues and suggests that further clarification is needed from DOD, as well as additional observation and research to be conducted.
The logistics threads of DOD supply chains will produce a new set of metrics and a new view of metrical units. How the military will manage the new, data-centric, and metric-centric logistics threads is still an unknown. There are still unknown unknowns about the total systems impact of passive RFID tags.
1. Hughey, Gary H. 2003. Military Door-to-Door Support to the Warfighter, US Transportation Command.
2. Hedgepeth, William Oliver. 2006. RFID Metrics: Decision Making Tools for Today's Supply Chains, Taylor & Francis Group/ CRC Press, UK.
3. Fiore, Amy Sommerfeld. 1986. Expert shell captures US Army regulations, Computerworld, Vol. XX, No. 35.
4. Dewar, James A., Carl H. Builder, William M. Hix, and Morlie H. Levin. 1993. Assumption-based Planning: A Planning Tool for Very Uncertain Times, RAND, Santa Monica, CA.
5. Wang, Mark Y. D. 2000. Accelerated Logistics: Streamlining the Army's Supply Chain, RAND.
6. RAND. 1995. Velocity Management: An Approach for Improving the Responsiveness and Efficiency of Army Logistics Processes. RAND.
7. Girardina, Kennethy J., William Lewis, Rick Eden, and Earl S. Gardner (1996). Establishing a Baseline and Reporting Performance for the Order and Ship Process, RAND, Santa Monica, CA.
8. Palfreman, Jon and Doron Swade. 1991. The Dream Machine: Exploring the Computer Age, BBC Books, London.
9. RAND Arroyo Center, Research Note: CWT and RWT Metrics Measure the Performance of the Army's Logistics Chain for Spare Parts, RB-3035-A, 2003.
10. Darilek, Richard, Walter Perry, Jerome Bracken, John Gordon, Brian Nichiporuk. 2001. Measures of Effectiveness for the Information-Age Army, RAND.
11. George, Michael L. 2002. Lean Six Sigma: Combining Six Sigma Quality with Lean Speed, McGraw-Hill.
ABOUT THE AUTHORS
Dr. Oliver Hedgepeth is the former Director of the Artificial Intelligence Center for Army Logistics and a retired civilian operations researcher for the US Army. He is currently the Associate Professor of Logistics and Chair of the Logistics Department at the University of Alaska Anchorage. He received his PhD in Engineering Management from Old Dominion University. He is author of RFID Metrics: Decision Making Tools for Today's Supply Chains. E-Mail: firstname.lastname@example.org.
Dr. Minnie Yi-Miin Yen is an Associate Professor of Management Information Systems at the University of Alaska Anchorage. She received her PhD in Management Information Systems from the University of Houston. Dr. Yen's work has appeared in IEEE Transactions on Software Engineering, Journal of Database Management, International Journal of Management, Human Factors in Information Systems, and a book chapter of Managing Business with Electronic Commerce: Issues and Trends. E-Mail: email@example.com.
by Dr. Oliver Hedgepeth, Associate Professor of Logistics, and Dr. Minnie Yen, Associate Professor of Management Information Systems University of Alaska Anchorage
RFID, metrics, barcode, transportation, logistics threads, assumption-based planning, RWT, CWT
RWT by Month 8/26/2006 1 2 3 4 5 6 Doc-Est 9748 22434 28884 8764 7615 17451 Est-MRO 5587 13046 17563 6441 4233 10524 MRO-Ship 4991 11041 9496 4116 3943 12571 Ship-Dra 4341 7260 8849 2280 1820 10380 EXHIBIT 3. Orders processed of RWT over a six-month period of time. Note: Table made from bar graph. RWT By Month Average Days /Segment 8/26/2006 1 2 3 4 5 6 DOC-EST 2.83 2.92 2.08 1.73 3.36 2.21 EST-MRO 5.61 4.06 4.12 3.31 4.82 3.46 MRO-SHIP 2.61 3.86 2.59 3.63 2.77 1.85 SHIP-DRA 7.71 12.68 15.82 11.75 9.48 10.58 EXHIBIT 4. Average processing time in days of RWT over six months. Note: Table made from line graph. Doc N DOC Date Requisition Proc. Date Z152526212XXXX 6212 6207 Z152526212XXXX 6212 6207 Z152526212XXXX 6212 6207 Z152526212XXXX 6212 6207 W91KGS6180XXXX 6180 6150 Doc N Requisition Proc. Date MRO Date W91KGP6047XXXX 6053 6048 HX2A062184XXXX 6215 6198 HX3A096184XXXX 6215 6199 FE50006024XXXX 6031 6027 W90DBX603XXXX 6059 6032 W91KGN6045XXXX 6053 6047 Doc N MRO Date Ship Rel. Date FB50006005XXXX 6074 6009 FB50006005XXXX 6046 6009 FB50046004XXXX 6092 6017 FB50006006XXXX 6074 6017 FB50006012XXXX 6094 6021 FB50006005XXXX 6081 6023 W91KGP6041XXXX 6076 6059 Doc N Ship Rel. Date MRA Rec. Date FB50005361XXXX 6005 6003 FB50005354XXXX 6009 6003 FB50006005XXXX 6093 6088 FB50005333XXXX 6030 6003 FB50046006XXXX 6010 6007 W81A3D5333XXXX 6003 6163 FE50005342XXXX 6034 6011 EXHIBIT 5. Sample of four data issues.
|Printer friendly Cite/link Email Feedback|
|Author:||Hedgepeth, Oliver; Yen, Minnie|
|Publication:||Defense Transportation Journal|
|Date:||Dec 1, 2006|
|Previous Article:||President's corner.|
|Next Article:||Thank you.|