Jan 1, 2000
How the Army Cut Order-and-Ship Time
For decades, through peace and war, a catalog of stubborn performance problems plagued the Army's order-and-ship process, the system Army personnel use to order and receive supplies. Each segment of the process—from placing a requisition for an item to receiving the package and every step in between—was not only slow, but also unreliable. Order-and-ship times (OST) for orders varied widely; some orders were delivered in a few days, but others took weeks, even when the ordered items were in stock. A lack of confidence in the reliability of the order-and-ship process led some Army personnel to hoard supplies and place duplicate orders. The Army long recognized these chronic problems, but repeated efforts to identify and eliminate their sources proved ineffective.
A new approach was needed to achieve change. In 1995, with the analytic support of RAND's Arroyo Center, the Army implemented the Velocity Management (VM) initiative, which adapts to the military many of the technological and managerial innovations that have proved successful in the commercial sector. Under the VM initiative, the Army has made impressive headway in improving the effectiveness and efficiency of the order-and-ship process and has begun to apply the approach to other logistics processes.
This Research Note summarizes the dramatic improvements in speed and reliability that have been achieved in the Army's order-and-ship process. It also explains how the Army accomplished this success and describes the many benefits that can flow from an improved order-and-ship process.
Most spare parts (Class IX supplies) needed by an Army installation are ordered from wholesale sources of supply, primarily the depots managed by the Defense Logistics Agency (DLA). Attacking the long and variable OSTs for these requisitions was the first focus of the Army's VM initiative. Figure 1 shows how OST has declined as a result. The bars represent the monthly OST performance for orders for repair parts (Class IX supply) placed by active units in the continental United States (CONUS) and filled by the wholesale supply system. Because this effort focused on improving the order-and-ship process for items on the shelf, backorders are excluded from the data. (Performance issues related to backorders are being addressed by a separate, though closely coordinated, process improvement team chartered to improve the stockage determination process.) The vertical dashed line distinguishes two periods of performance. The period from July 1995 through December 1997 represents performance trends since the Velocity Management initiative took hold. The twelve preceding months, July 1994 to June 1995, are the baseline period and serve as the basis of comparison for gauging progress. The segments on each bar measure the month's OST performance at the 50th, 75th, and 95th percentiles. (The 50th percentile indicates the day by which 50 percent of the orders are filled, the 75th 75 percent, and so on.) The line running through all bars is the average OST.
As shown by the continuing downward slope of the bars and line, the Army has made dramatic and nearly continuous improvements in the order-and-ship process under Velocity Management. The performance during the baseline period for the three percentiles was 17, 25, and 56 days for the 50th, 75th, and 95th percentiles respectively, with an average OST of just over 22 days. Corresponding figures for December 1997 were 8, 13, and 25 days, with an average OST of under 11 days—in short, approximately a 50 percent reduction at all percentiles.
Compared to CONUS units overall, some of the large Forces Command installations that were among the first to participate in the Velocity Management initiative have achieved even greater gains, suggesting that other units can also expect to achieve further reductions. Figure 2 shows the results for active units at Fort Bragg. There the median OST has declined from a baseline average of 18 days to 5—a 72 percent reduction. The 75th and 95th percentiles show similar improvement. Average OST has fallen from 26.3 days for the baseline period to 7.3 for December 1997.
Improvement has also occurred in order-and-ship process performance for overseas units ordering parts from CONUS depots. Median OST for units outside CONUS has declined from 23 to 18 days, a 22 percent improvement. The gains in the other percentiles have been even greater. Outside CONUS units now receive 75 percent of their orders in 26 days (a 47 percent improvement) and 95 percent in 55 days (a 35 percent improvement). That these reductions are less, thus far, than those achieved by CONUS units reflects both the additional complexity of the OCONUS distribution system and the fact that VM was generally implemented later by overseas installations.
Possibly the most important aspect of the Velocity Management approach to achieving needed change is a high-level commitment under the VM initiative to improve the process end-to-end, including those segments that fall outside the Army's direct control. This focus is essential, since functional and organization boundaries had caused a myopic focus on segments of the process, sometimes at the cost of overall effectiveness and efficiency. The commitment to change was endorsed by the Army's senior leadership and is overseen by a coalition of senior logisticians (the Velocity Group), led by the Deputy Chief of Staff for Logistics, the Deputy Commanding General of the Army Material Command, and the Commanding General of Combined Arms Support Command. The Velocity Group also includes senior representatives from Forces Command, U.S. Army Europe, U.S. Forces Korea, DLA, and the U.S. Transportation Command.
The Velocity Group organized cross-functional teams focused on improving specific logistics processes such as order and ship, repair, stockage determination, and financial management and staffed them with technical experts from across the Army and, when appropriate, outside the Army as well. These process improvement teams (PITs) are complemented at each installation by Site Improvement Teams (SITs) that are similarly composed of personnel representing all functions and segments of the targeted process.
The improvement teams employ a define- measure-improve (DMI) methodology to build the collective expertise and coordination necessary to achieve dramatic and continuing improvement.
"Define," the first step, aims at producing a clear picture of the entire process that the team is attempting to improve. In the case of improving the order-and-ship process, it involved experts from many organizations, both from the Army and elsewhere, literally walking the process from the time the need for a part is identified until the part arrives in the hands of the mechanic who is going to install it on the equipment. Such a detailed definition phase is critical because though many personnel are expert in their own segment of the process, no one has a detailed understanding of every stage. A variety of functions are involved in the order-and-ship process—transportation, maintenance, supply—and improving the process requires the involvement of all functional players.
Once the process was defined, it was necessary to determine the best way to measure it to foster improvement. The metrics selected were time, quality, and cost. Focusing first on the time it takes from placing an order to receiving the item, measurement in terms of the percentiles that appear on the charts provided a useful picture of the goodness of the process. These metrics were a marked departure from the traditional measure of average OST. The percentiles gave information on typical OST performance and focused efforts on reducing the wide variations in delivery time associated with the orders that take the longest time to be filled and delivered.
The order-and-ship PIT used the data associated with the metrics to help diagnose systemic problems. Statistical analysis and data mining were used to identify sources of delay. Another methodology used successfully by the SITs was a report that listed each requisition whose OST lay beyond the 95th percentile. These "outliers" were researched individually by personnel operating in each segment of the process to identify and eliminate the sources of such extraordinary delay.
The definition and measurement stages showed that many segments of the process were being managed with metrics that did not necessarily result in good customer service. For example, some organizations and the segments managed by those organizations measured themselves by the efficient use of trucks, so partial truck loads were held up until a full one could be assembled. While this goal and this metric yielded more efficient use of trucks, it delayed getting the needed part to the customer and lengthened OST for many orders. Other examples exist of conflicting goals that resulted in the apparent efficient use of some resources at the overall expense of the whole.
The final stage of DMI, "improve," involves combining the end-to-end understanding of the process developed in the "define" stage with the diagnoses of the sources of performance deficits that were isolated in the "measure" stage. Once likely process improvements were identified, the Army implemented the changes that it could do on its own. Army installations strengthened oversight, simplified rules, improved the use of new requisitioning and receipting technologies, reduced review processes, streamlined on-post delivery, and made use of the information available from the new metrics. Other changes required establishing partnerships with the organizations that controlled other segments of the order-and-ship process, such as DLA, which controls the depots, and commercial trucking and small package delivery firms. For its part, DLA has improved workflows through its distribution depots, sped up the processing of Materiel Release Orders, packaged and directed shipments to reduce intermediate handling on post, and worked with commercial shippers to provide scheduled deliveries.
The analyses of order-and-ship process performance that the Arroyo Center conducted in support of the Army's VM initiatives included diagnoses of two segments of the order-and-ship process that were not under the Army's direct control—namely, the processing of orders in the depot and the movement of items from the depot to the installations. These analyses found that much of the delay and variable shipping times in these segments reflected DLA's practice of using a variety of shipping modes in an attempt to match each order with the lowest-cost shipping mode that was appropriate to its urgency and characteristics such as size and weight. The mixing of modes caused some orders to be delayed (for instance, to wait until enough similar orders accumulated to fill a truck) and required the installations receiving the materiel to cope with multiple deliveries, most of them unscheduled.
The analyses suggested strongly that DLA could greatly reduce the delays and variability in the depot and transit segments if it could establish scheduled trucks (similar to regular mail deliveries) as the primary shipping mode to Army installations and then synchronize other activities, particularly depot processing, with these regular delivery schedules. Many depot-post combinations had driving times of one day (in most cases two days or less) and sufficient volume to support trucks daily or every other day. In such cases, even high-priority items that were formerly shipped by air could be placed on these trucks, saving the expense of premium transportation services such as FedEx.
Working closely with the Army, DLA agreed to increase the use of scheduled truck shipments for large installations. Under the scheduled truck concept, depots that serve large installations place all the shipments for that installation, regardless of eligibility for air shipment or bulk considerations, on a routinely scheduled truck. To increase the opportunities for capitalizing on scheduled trucks, DLA has had to implement a number of changes. Its depots have applied automation to sort shipments into multipacks and have added automated manifest cards for key customers on post, reducing workloads and decreasing the time required to receipt shipments. The stock positions at some depots are being changed to reflect the needs of the closest Army installations. This will permit more volume to flow between these depot-installation combinations at no additional transportation cost; in some cases it will increase the number of trucks that it is cost-effective to send per week. More frequent deliveries means lower OST.
The Army and DLA are making quick and continuous progress at increasing the use of scheduled trucks between major CONUS installations and their primary depots. By the end of 1997, between 40 and 50 percent of all shipments from DLA depots to major Army installations in CONUS were sent on scheduled trucks, and the mean OST for these shipments was four to five days faster than the mean OST for other shipments combined. In December 1997, for instance, shipments on scheduled trucks averaged just 9.3 days OST, whereas other shipments averaged 13.5 days.
As Figure 3 indicates, an improved order-and-ship process has many benefits, both direct and indirect. Most directly, improved OST means the quick and dependable delivery of spare parts throughout the Army.
In addition to giving customers what they need when they need it, a quick and reliable order-and-ship process also reduces the number of orders in the system because Army personnel are no longer frustrated into reordering a delayed part several times. They also have less incentive to hoard parts because they are more confident they will get them when they order them.
Improved OST also improves the Army's repair and stockage-determination processes. The repair process benefits because less time is lost awaiting parts, a major source of delay in today's repair process. Local stockage improves because fast delivery means stocks do not have to be as deep. Some of the money saved from having shallower stocks of a given item can be reinvested in providing a broader array of parts. Thus, more needed parts are available locally, further speeding repairs.
The VM improvements in the order-and-ship process have been impressive, but the VM initiative is not limited only to the order-and-ship process nor only to reducing cycle times. The Velocity Group has directed other improvement teams to apply the Define-Measure-Improve methodology to the repair process and the stockage-determination process as well as to the financial management process. Moreover, for each of these processes, the goal of the PITs is to identify and eliminate sources not only of delays but also of errors and waste. Focusing on time, quality, and cost will deliver a logistics system that is faster, better, and cheaper.