Improving the logistics pipeline: achieving Agile Combat supply support - Logistics Support

Steven L. Martinez

A pipeline on the fly concept, deploying smaller spares packages and leveraging immediate resupply from a centralized location, can enable supply support to become light, lean, and lethal.

From an Air Force perspective it is imperative that the logistics footprint or support personnel and equipment required by an aerospace expeditionary for (AEF) be reduced. The goal is to "streamline what we taken with us, reducing our forward support footprint by 50 percent." (1) By doing this, units can deploy much more quickly, and critical lift forces (usually airlift) required to move them can be used for only the most urgent requirements. The popular catchphrase to describe this characteristic is light lean, and lethal.

In fact, the goal of the Air Force is to be able to deploy an AEF within 48 hours and up to five AEFs in 15 days. This will be done through improvements generated by leveraging "information technology, rapid transportation, and the strengths of both the organic and industrial logistics base to ensure responsive, dependable, precise support." (2)

Within the realm of supply support, the movement of spare parts and key consumable items, normally contained in a readiness spares package, is a major consideration for planning the deployment of a unit. As such, methods to reduce the size of mobility readiness spares packages (MRSP) must be investigated. Currently, MRSP requirements are computed based on 30 days of support for a contingency, with the assumption there will be no resupply. The amount of spares authorizations allotted to each base for every weapon system comprises the assets needed to support the most taxing scenario involving the greatest number of aircraft that would deploy from that location. In practice, supply and sortie generation personnel coordinate with each other to tailor each kit, based on the expected number of sorties and duration of each sortie for the contingency. However, it seems as though there is no situation, except for a deployment that you cannot resupply within 30 days, for which it is necessary to keep 30 days of spares on hand. Therefore, it seems logical, for cost and airlift-requirement reduction purposes, to stock at the home station only the minimum number of spares required to support a deployment, up to the point at which the resupply pipeline can deliver an asset to the forward operating location.

Also, it is probable that the Defense Transportation System, through which aircraft parts are moved, can be improved so that the assets needed for an entire military operation do not have to be deployed at the outset of a contingency. In contrast, by reducing the total shipment time and the variability in these times, holding at the home station those spares projected to be needed later in the deployment may be a viable way to reduce the initial lift requirement (Figure 1).

The parts needed after the first few days of the conflict could be shipped from the depot at the same time as the deployment from home station, and those parts would be available as the spares from the kit began to deplete. This concept is known as the pipeline on the fly. An added benefit from this technique is that parts flowed to the forward operating location later in the contingency would be only items specifically requested by the deployed unit, rather than continuing to be comprised of parts estimated to be needed in the deliberate planning process (Figure 2).

The Air Force could maintain smaller spare parts kits and hold some of the assets no longer stored in the base-level MRSPs at a higher echelon inventory point--centralizing the inventory. This would allow a lower overall level of inventory, Air Force-wide, to attain the same service level as can be achieved with the current decentralized spare parts kits.

Risk Pooling

The concept of risk pooling demonstrates the benefits that can be derived from transforming an inventory system from a decentralized structure to a more centralized network.

Risk pooling suggests that demand variability is reduced if one aggregates demand across locations because, as we aggregate demand across different locations, it becomes more likely that high demand from one customer will be offset by low demand from another. This reduction in variability allows us to reduce safety stock and, therefore, reduce average inventory. (3)

Tradeoff Framework

While there are numerous considerations involved when incurring costs in a business enterprise, there must exist a method by which these costs are categorized and compared. A classic methodology used in the study of logistics to decide on the implementation of just in time (JIT) is to compare costs using the inventory-transportation tradeoff (Figure 3). In this type analysis, it is given that a firm desires to decrease total operating costs and is weighing decreases in on-hand inventory costs (for example, purchasing, warehousing, and personnel) from carrying less material against increases from using premium transportation to move items quickly and consistently through the logistics network. It must be stressed that it is much more important to reduce the variability of the transportation than it is to speed it up. A process that is reliable and consistent allows for more effective planning and forecasting of demands.

Logistics Response Time

To analyze the logistics pipeline, there must be a useful way to measure it. The process of transforming a need into an asset in hand recently has been evaluated in several ways. The following is a discussion of the two most recent measurements that have been used by the Air Force: order and ship time (O&ST) and logistics response time (LRT).

It is important to understand how the Air Force Materiel Command (AFMC) LRT metric ties into the logistics pipeline model. Each segment can be aligned with a step in the pipeline, as shown in Figure 4. The base requisition time reflects step 1, the time required to transmit an order from the requester to the source of supply. In step 2, both the ICP (order receipt) and Defense Logistics Agency (DLA) (acquisition or order picking) processes occur at the depot. The transit time reflects the rest of the logistics pipeline, from the time the depot inputs an item into the transportation system until the item is received by the user and status is updated to reflect the asset arrived.

The time it takes to place an order for an item from the forward operating location and receive it had to be calculated. This provided the frame of reference for determining what range of resupply times is probable in future contingencies. Sample data taken from Operation Noble Anvil, the US air campaign in support of Operation Allied Force, and the US and North Atlantic Treaty Organization's action to bring an end to Serbian atrocities in Kosovo were analyzed statistically to construct the feasible region of times.

The data tended to follow a lognormal distribution, as ascertained through the use of a distribution analysis software program. Because the values follow such a distribution (Figure 5), it is more valid to view the median or mode as a measure of central tendency than the mean or average logistics response time.

The median is "the middle number when the measurements are arranged in ascending (descending) order." (5) Another way to describe the significance of this statistic is to note that 50 percent of the area under a graph of the distribution of values lies to the left of the median, and 50 percent of the area lies to the right. This statistic is a more valid measure of central tendency than the mean since it is less susceptible to the effects of very large or very small data values. In addition, the mode was considered in the research since it "is the measurement that occurs most frequently in the data set." (6) This statistic is especially useful in cases when it is important to ascertain the section of the quantitative data set in which most of the observations occur. (7) As shown previously in Figure 5, the skewing of the data results in a mean value that is much higher than the median. So consideration of the median and mode was appropriate (Table 1).

The results shown in Table 1 indicate that the current logistics pipeline--tested in one of the most recent combat situations--performs rather well, since a part almost always arrived where needed in 6 days. However, it seemed that the process included a lot of variance and, hence, made it less than reliable. Compared to the descriptive statistics for the O&ST values used in calculating the kit spare parts requirements, the current pipeline seemed to perform better.

A more indepth analysis of the logistics response times was accomplished by identifying quantiles within the original distribution and eliminating values that occurred in the highest sections. These occurrences are known as outliers and, typically, are anomalies or random errors that can be found in any process. By removing these values that may not be representative of the true performance of the system, one can gain better insight into the factors influencing its operation.

The three outputs in Table 2 represent the elimination of the highest 5 percent (95), 10 percent (90), and 25 percent (75) of the LRT values, respectively.

As the data indicate, perhaps a better approximation of the actual average logistics response time is around 16-18 days.

Impact of Kit Reductions

Given a specified direct support objective, the impact of the reduction of MRSP sizes to satisfy demands until resupply is established was studied. Specifically, the amount of spares investment cost and airlift requirement that could be eliminated by assuming the logistics pipeline could react more quickly than currently possible was calculated.

Aircraft Sustainability Model

The Aircraft Sustainability Model (ASM) is used by the Air Force to calculate the number of spares required to be maintained in an MRSP. The logic of the program ensures that the spares mix producing the highest aircraft availability, given a level of funds, is created. The model requires data elements provided by either the Dyna-Metric Microcomputer Analysis System or the D087 report from AFMC, known also as the Requirements Execution Availability Logistics Module (REALM). REALM contains information pertaining to items such as demands (failures) per flying hour, base and depot repair times, probability of repair at a given location, condemnation rates, shipping times, unit cost, quantity per application, and procurement lead time. (8)

Once these data are imported into the model, the program initiates a three-step process as described below:

* The first step involves characterizing the probability distribution of the number of items in various stages of the resupply process (or pipeline)--unserviceables in repair at bases or depot and serviceables and unserviceables in transit. The relationship between these quantities and the number and location of spares in the system determines the probability of a back order.

* The second step is to relate that item information to weapon-system performance; specifically, to determine the expected number of item back orders, the expected number of aircraft not mission capable supply, and several other weapon-system-oriented measures of supply performance.

* The third step is to produce the availability-versus-cost curve and the associated optimal spares mix for a specified availability or budget target. The model uses a marginal analysis technique that determines the best mixes of spares for a wide range of targets. (9)

This technique is illustrated in Figure 6. In the first step, the user inputs information, based on either a steady-state (peacetime) or dynamic (wartime) flying-hour scenario, into the model. Since the research analyzed support of combat operations, dynamic flying-hour data were used. The second step actually calculates the expected aircraft availability based on the cannibalization option chosen in step 1. Then the third step allots an optimal spares mix through marginal analysis, recommending the purchase of items that have the highest benefit-to-cost ratio first.

It creates a shopping list of spares and purchases, each one in order, until either all spares are allocated or the specified funding level for spares is exhausted.11 Because of the importance of generating every sortie in wartime operations, cannibalization or the removal of a working item from a nonfunctional aircraft to another aircraft is a normal practice. Therefore, the full cannibalization option was used throughout the research.

To evaluate the effect of changes to the logistics response time on spares requirements, it was necessary to adjust the data within the kit files from D087 to reflect various average order and ship times. Although logistics response time includes more than just order and ship time, the only other repair time values in the kit file were base repair and depot repair. There was no point in considering base repair time since it was assumed that none would be available. This assumption is discussed later when the values for base repair time of components is explained. Also, depot repair time values include more data than does the logistics response time. Therefore, including this number in the analysis could have injected more error. So the simplest and most accurate substitute for logistics response time was order and ship time.

The adjustment of O&ST values was accomplished by exporting the kit data into an Excel spreadsheet and modifying the values representing expected wartime order and ship times for each item in the kit (Table 3).

Using the solver add-in, these values were adjusted to provide overall average order and ship times of 5, 10, 15, and 20 days for the entire kit of each aircraft type. In Table 3, the % Reduction column represents the percentage decrease applied to the original values that result in an average (Average) that is equal to or less than the target (Goal) value. The spares packages for all four aircraft had average order and ship times less than 25 days, so there was no need to create higher adjusted average values. These new item order and ship times had to be rounded to the nearest integer, put back into the Excel spreadsheet, and input into the Aircraft Sustainability Model.

Airlift and Cost Savings As a Result of a More Rapid Logistics Pipeline

Experimental data run in the Aircraft Sustainability Model were accomplished for each aircraft--B-52H, F-15E, F-16C, and KC135--with various combinations of order and ship time and the day order and ship begins (DO&SB) (Table 4). For each weapon system, the number of aircraft the kit was designed to support (primary aircraft authorized) was matched with various values of order and ship time and DO&SB.

Similar combinations of values were used for each of the weapon systems in this analysis, with the value for primary aircraft authorized based on the size of actual spares kits used in the Air Force. Once the total cost of the kit was calculated, it was compared with the cost of the current 30-day kit. A percentage difference was computed to show the degree of decrease that results from the changes in order and ship time and DO&SB. An example of the results attained in this analysis is shown in Table 5.

These analyses indicate there may be significant cost savings that can be achieved by either reducing the order and ship time or DO&SB, or both. Further, these reductions can be attained while still maintaining the minimum level of support (target aircraft availability rate) used to compute spares requirements.

Just as compelling were the reductions in the kit size realized through the changes in order and ship time and DO&SB. Again, a sample of the reductions in kit size is shown in Table 6. As was seen in the values for kit cost, there were significant reductions in kit size when the order and ship time and DO&SB were decreased.

The Effect of Pipeline on the Fly

Last, a determination had to be made as to whether the pipeline on the fly approach would yield any significant reduction in the MRSP requirements. By modeling the effects of this adjustment to the current process, the resultant improvement was calculated and analyzed for its significance.

The response surface graph in Figure 6 is an example of the illustrations created to give an indication of the relative strengths of both independent variables in producing the value for kit cost and kit size. For the B-52H, there was a distinct linear decrease that corresponded with the decrease in order and ship time. Also, there was almost no variation in the axis that represents the values for DO&SB.

In fact, the response surfaces for the F-15E, F-16C, and KC-135 indicated the same relationships. All showed a decline in the kit cost and size values that match the trend in order and ship time, and very little changed in relation to the decrease in the variable DO&SB. Visually, it was apparent that order and ship time had a significant impact on the kit cost, while it seemed that DO&SB had very little influence on the reductions that occurred. The response surfaces for kit size also illustrated this relationship, an example of which can be seen in Figure 7.

Regression analyses were accomplished to better understand the effects of the two independent variables: order and ship time and DO&SB. The null hypothesis for this experiment was that there was no difference in the coefficients of all regression terms, while the alternate hypothesis was that there was at least one regression coefficient that was different.

For the equation

y = [[beta].sub.0] + [[beta].sub.1][x.sub.1] + [[beta].sub.2][x.sub.2] + [epsilon],

where y = kit cost/kit size, [x.sub.1] = O&ST, and [x.sub.2] = DO&SB, then

[H.sub.0]: [[beta].sub.1] = [[beta].sub.2] = 0

[H.sub.[alpha]]: [[beta].sub.1] [not equal to] 0 or [[beta].sub.2] [not equal to] 0

The results (Table 7) indicated that the variable order and ship time was the only significant contributor to the value of kit cost for F-15Es.

The change in DO&SB did not have a significant impact on the dependent variable kit cost. The same was true for all kit sizes and costs, except for the B-52H (both kit size and cost) and the KC-135 (kit size). It can be concluded, then, when only these two variables were considered in a model, order and ship time was a significant predictor of the output results while the effect of DO&SB was not clear.

FSL Option of the Aircraft Sustainability Model

To better evaluate the pipeline on the fly concept, the forward support location (FSL) option of the Aircraft Sustainability Model was employed. In its basic form, the FSL option allows the user to analyze the spares level required when using a central inventory point that is in the same geographic area as the spare parts kits at the forward operating locations. The forward support location, then, is an intermediate storage location between the end user and the depot. In the research, the objective was to understand the feasibility of the pipeline on the fly concept. Therefore, the FSL option was used in a modified manner so that it would model only the stockage of aircraft spares at depots and forward operating locations (Figure 8).

By setting the resupply time from the depot parameter at a value of 99, the model effectively stocked at only two echelons: the forward support location and forward operating location. Assuming there was a requirement to stock an asset at either the forward support location, with a reasonable order and ship time, or the depot, with an order and ship time of 99 days, the model always chose to place it at the forward support location. Henceforth, the forward support location can be thought of as the depot, and the depot can be thought of as the manufacturer. When this was done, the model was, in effect, forced to stock an asset either at the depot or at one of the forward operating locations.

When the FSL option was used, it seemed the pipeline on the fly concept was modeled more aptly. In contrast to the use of DO&SB in the Aircraft Sustainability Model, the FSL option provided results that could be used to illustrate the impact on the logistics pipeline from implementing a change in the process. An example of the results obtained from the FSL option is shown in Table 8.

In Table 8, Percent Reduction was the difference between the 30-day value (either cost or size) and the value obtained at the various order and ship times. Then, Kit Sum Cost/Size were the individual kit sizes or costs multiplied by the number of spare parts kits in the Air Force. The Depot Cost/Size represented the amount of spares that the FSL option recommended for stockage at the depot and, when added to the Kit Sum Cost/Size, became the Overall Cost/Size. Finally, the 30-Day Kit Cost/Size reflected the cost and size of a standard spares kit analyzed in the Aircraft Sustainability Model with the same sortie data (number of sorties per aircraft, hours per sortie, and total hours per day), and that standard kit multiplied by the number of kits is the overall 30-Day Kit Cost/Size.

At this point, using a graphical depiction of the results helps one appreciate the magnitude of savings possible by using this type analysis. The Air Force's newest airlifter, the C-17 Globemaster III, is capable of carrying a maximum payload equal to 18 pallets.(12) Assuming a typical AEF deployment consists of at least one MRSP from each of the four weapon systems analyzed in the research, such a movement would require 66 pallets of parts and cost $85,510,862.33. (13) To move this load, the Air Force would need to use 3.67 C-17 aircraft (Figure 9).

In contrast, simply using the FSL option with no reduction in order and ship time (O&ST = Baselines) lowered the single-deployment airlift requirement by nearly 24 percent. The airlift requirement gradually slimmed to .59 C-17s when the order and ship time was cut to 5 days. When the size of each spares package was multiplied by the number of kits the Air Force maintains and added to the size of spares stocked at the depot, an overall kit size was the result (Figure 10).

Again, the current 30-day kits, when analyzed with the FSL option, were immediately reduced by almost 27 percent to 418.68 pallets. The amount of spares continued to decline until it was the equivalent of 4.71 C-17 loads when the order and ship time was 5 days, an 85-percent reduction from the current kit levels.

While these results are significant with respect to the Air Force's objective of reducing its deployment footprint, the cost savings attained through the use of FSL option analyses are perhaps more amazing. When compared with the cost of a single deployment of current 30-day kits, using the FSL option, without adjusting the order and ship time, lowered the cost by more than 28 percent to $61,279,584.88, for a savings of $24,231,584.45 (Figure 11).

The savings achieved by using the FSL option and reducing the order and ship time to 5 days nearly equaled the cost of a single deployment of current 30-day MRSPs.

In a fashion similar to overall kit size, the cost of each kit was multiplied by the number of kits the Air Force had and added to the spares stocked at the depot to calculate an overall kit cost. Again, merely using the FSL option with the baseline kit data resulted in almost a 27-percent reduction in the cost of aircraft spares, from $714,862,875.61 to $512,163,811.78. This savings was the same amount needed to purchase almost one C-17 aircraft (Figure 12).

Lowering the order and ship time to 5 days further increased the savings to $628,719,490.99 or the cost of 2.66 C-17 aircraft.

Summary

The research effort was conducted to gain an understanding of the effect improving the logistics pipeline had on the way the Air Force supplies aircraft spares in combat operations. Through various improvement efforts, the Air Force is attempting to streamline its logistics functions. This will enable a future AEF to be employed as a light, lean, and lethal combat power. Two main areas in this endeavor are reducing the cost of support and trimming down the size of the materiel needed for this support.

Assessed in the research were the effects of reducing the logistics response time and implementing a pipeline on the fly technique. To fully comprehend the impact of both these efforts, the research was structured to answer five main investigative questions:

* What is the logistics pipeline?

* How quickly can the logistics pipeline be established?

* How long does it take to place an order and receive a part in the logistics pipeline?

* How much airlift and funding can be saved by reducing kits to support an operation when a logistics pipeline that can respond more quickly than currently possible exists?

* Does the pipeline on the fly concept yield a significant improvement in logistics pipeline performance?

The remainder of this article will answer these questions, discuss any conclusions that can be drawn from this analysis, and recommend research that would continue to add insight into this area of logistics.

What is the logistics pipeline?

The logistics process has been described as a pipeline for many years, but the specific methods used to measure it have been adjusted several times. Today, it encompasses the entire order cycle, from identifying the need to satisfying that need. It begins with the input of a requisition for a particular item by a specific unit, now mostly done through an online computer system. Then that order is transmitted to the respective source of supply, where it is analyzed and processed. Once an asset is available to fulfill that requirement, it is shipped to the requesting organization.

A measurement that is currently used by the Department of Defense and the Air Force is the logistics response time. To date, it is the metric most representative of the various segments comprising the logistics pipeline. As such, the logistics response time is the key concept around which the research was conducted, and its reduction and its effect on MRSPs was one of the main objectives.

How quickly can the logistics pipeline be established?

While the assumption was made that base support for an AEF would be in place within 48 hours after the deployment order is given, the literature pointed to several issues that have kept the goal from becoming a reality. However, the example cited was only a sample of one event that occurred in l997. (14) Therefore, it is likely that, through various subsequent exercises and simulations, functions required to enable supply processes at a bare base could now be in place earlier than 1 week after the deployment commences. Just what that number of days is cannot be ascertained at this point. However, it seems safe to conclude that it would occur much earlier than the minimum reasonable order and ship time, now or in the near future. Even if a part was shipped on day 0 of a contingency, the order and ship time required to move that asset to the forward operating location would exceed the number of days needed to set up supply operations. Therefore, it does not seem that the time needed to make a deployed location fully operational would be of great concern in relation to the time required to establish a viable logistics pipeline.

How long does it take to place an order and receive a part in the logistics pipeline?

Based on LRT data collected from Operation Noble Anvil, the mean time to order and receive a part was 36 days. However, the distribution of times was not normal; rather; it was best modeled by a lognormal distribution. Therefore, more valid indicators of the central tendency of the logistics response times were the median and the mode. These values were 15 and 6 days, respectively. Therefore, it was highly probable during this contingency that an asset would require from 1 to 2 weeks for delivery. As such, the current logistics pipeline is not very far from being able to perform well enough to produce average order and ship times like those used in the research. It may not require much more effort or resources to achieve an average order and ship time of 10 or even 5 days, since the pipeline most often can move assets within times ranging from 6 to 15 days.

These promising data probably came about because of increased attention and focused management, for they represented materiel being moved in support of actual combat operations. So it can be presumed that any future conflict will enjoy a similar level of support from all agencies and functions comprising the logistics pipeline. Such an assumption may not be prudent from a military planning standpoint though.

How much airlift and funding can be saved by reducing kits to support a logistics pipeline that can respond more quickly than currently possible?

The experiments conducted using the Aircraft Sustainability Model resulted in tremendous cost and size savings for MRSPs when both order and ship time and DO&SBs were reduced. All weapon systems considered--B-52H, F-15E, F-16C, and KG135--experienced reductions in both cost and size from approximately 4 to 90 percent and more. In fact, when the average order and ship time is 5 days, the model recommended no kit at all for the KC-135. Clearly, there is much to be gained, both in saving scarce funding and minimizing the logistics footprint when deploying forces, by endeavoring to reduce order and ship time and DO&SB. Again, these results were not exact since notional sortie data had to be used. However, they did give an indication of the magnitude of savings that could be achieved by improving the logistics pipeline.

On a particular deployment, units already reduce their spares kits (paring and tailoring) to take only those items required for a specific scenario. The savings described here would be obtained by decreasing the number of spares kept on hand on a day-today basis, since we would not be stocking with the 30-day, no-resupply assumption for every weapon system at every base. However, a key question remained as to which variable would produce the more significant reductions in kit sizes and costs.

Does the pipeline on the fly concept yield a significant improvement in logistics pipeline performance?

Based on the regression analysis conducted to determine the significance of order and ship time and DO&SB on the value of the independent variables kit cost and kit size, it was evident that DO&SB was almost insignificant. The resultant values of kit cost and size were affected almost completely by order and ship time. By this result alone, it seems that efforts to reduce the cost and the size of Air Force MRSP should focus on ways to reduce order and ship time rather than DO&SB. However, the results obtained through the use of the FSL option of the Aircraft Sustainability Model indicated there might be significant benefits--namely, savings in cost and airlift requirement-that could be achieved through the implementation of the pipeline on the fly technique. In fact, the unique adaptation of the FSL option created during the research pointed to the possibility that the Air Force could save more than 80 percent in both spares cost and cargo movement needs when the pipeline on the fly approach is combined with a reduction of the order and ship time to 5 days.

[FIGURE 3 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

Table 1

Excel Descriptive Statistics Output for Allied Force and Noble Anvil
Logistics Response Times

Operations Alllied Force and Noble Anvil LRTs

Mean                      36.41511224
Standard Error            1.050801383
Median                             15
Mode                                6
Standard Deviation        59.09765275
Sample Variance           3492.532561
Kurtosis                  22.57363628
Skewing                   3.674828978
Range                             698
Minimum                             1
Maximum                           699
Sum                            124670
Count                            3163
Confidence Level (95.0%)  2.060323046

Table 2

Descriptive Statistics without Outliers

                             95%          90%          75%

Mean                     29.1464226   24.18042494  14.16814159
Standard Error           0.61789871  0.471770.283  0.212868542
Median                           14            13           11
Mode                              6             6            6
Sandard Deviation        33.8718977   25.27826007  10.36956061
Sample Variance          1147.30545    638.990432  107.5277873
Kurtosis                 3.02132817   2.275747393  0.570876308
Skewing                  1.89318719   1.708228682  1.181650125
Range                           161           112           45
Minimum                           1             1            1
Maximum                         162           113           46
Sum                           87585         69422        33621
Count                          3005          2871         2373
Confidence Leve (95.0%)  1.21154635   0.925041851  0.417427989

Table 3

Sample Solver Calculations

                                                               %
NSN               78.4   56.8    35.2    13.6     0.0     0.0  Reduction

1560007242853FL     13     19      26      30      30      30
1560008601911FL     13     19      26      30      30      30
1560008601912FL     13     19      26      30      30      30
1560011273340FL      6      8      11      13      13      13
1560010639477       13     19      26      30      30      30
1630004927144       13     19      26      30      30      30
        ///        ///    ///     ///     ///     ///     ///
6620005573023       13     19      26      30      30      30
6620005619380        7     11      15      17      17      17
6620011404405        7     11      15      17      17      17
6620011450265       13     19      26      30      17      17
6620011519590       13     19      26      30      14      14
6620012471816       13     19      26      30      30      30
Average          4.806  9.982  14.783  20.083  23.138  23.138
Goal                 5     10      15      20      25      30

Table 4

Sample of Experimental Runs in the Aircraft Sustainability Model

A/C    PAA  O&ST  DO&SB

B-52H   6     5     0
B-52H   6     5     7
B-52H   6     5    15
B-52H   6    10     0
B-52H   6    10     7
B-52H   6    10    15
B-52H   6    15     0
B-52H   6    15     7
B-52H   6    15    15
B-52H   6    20     0
B-52H   6    20     7
B-52H   6    20    15

Table 5

Sample of Results from ASM

                         Kit Cost
 A/C   PAA  O&ST  DO&SB    ($M)    % Diff *

B-52H   6     5     0     15.56     61.55
B-52H   6     5     7     16.65     58.86
B-52H   6     5    15     16.65     58.86
B-52H   6    10     0     23.57     41.76
B-52H   6    10     7     24.19     40.23
B-52H   6    10    15     24.71     38.95
B-52H   6    15     0     30.58     24.44
B-52H   6    15     7     30.92     23.60
B-52H   6    15    15     32.10     20.69
B-52H   6    20     0     37.35      7.71
B-52H   6    20     7     37.68      6.90
B-52H   6    20    15     38.80      4.13

* vs 30-day kit cost of $40.47M

Table 6

Sample of Results from ASM Experimental Runs, B-52H Kit Size

                         Kit Size
 A/C   PAA  O&ST  DO&SB  (Pallets)  %Diff *

B-52H   6     5     0       7.84     67.05
B-52H   6     5     7       9.95     58.20
B-52H   6     5    15       9.95     58.20
B-52H   6    10     0      12.33     48.19
B-52H   6    10     7      13.10     44.98
B-52H   6    10    15      13.57     42.98
B-52H   6    15     0      14.74     38.07
B-52H   6    15     7      14.81     37.76
B-52H   6    15    15      16.26     31.69
B-52H   6    20     0      17.75     25.42
B-52H   6    20     7      17.88     24.86
B-52H   6    20    15      20.59     13.48

* vs 30-day kit size of 28.8 pallets

Table 7

Sample Regression Analysis Results from JMP, F-15E Kit Cost

Response: F-15E Kit Cost Summary of Fit

R Square                    0.958523
R Square Adj                0.949306
Root Mean Square Error      0.204555
Mean of Response            0.204555
Observations (or Sum Wgts)  0.967512

Effect Test

Source  Nparm  DF  Sum of Squares  F Ratio   Prob>F

O&ST      1    1     8.70220417    207.9702   <.001
DO&SB     1    1      0.0007988      0.0191  0.8931

Table 8

Sample FSL Option Results, F-15E

F-15E    No of Kits           3
O&ST      Kit Cost       % Reduction     Kit Sum Cost

 5        $349,725.54       97.49        $1,049,176.62
10      $2,744,519.97       80.28        $8,233,559.91
15      $5,011,562.15       63.99       $15,034,686.45
20      $7,430,842.24       46.61       $22,292,526.72
21      $7,787,953.55       44.04       $23,363,860.65

       30-Day Kit Cost  $13,917,843.06

F-15E
O&ST         Depot Cost          Overall Cost      Kit Size

 5          $3,286,395.21        $4,335,571.83       0.23
10          $3,286,395.21       $11,519,955.12       1.20
15          $3,286,395.21       $18,321,081.66       1.71
20          $3,286,395.21       $25,578,921.93       3.33
21          $3,306,465.48       $26,970,326.13       3.43

       Overall 30-day Kit Cost  $41,753,529.18  30-Day Kit Size

F-15E
O&ST   % Reduction  Kit Sum Size        Depot Size         Overall Size

 5        94.18         0.68               1.70                2.37
10        68.96         3.60               1.70                5.30
15        55.74         5.14               1.70                6.84
20        14.00         9.99               1.70               11.68
21        11.39        10.29               1.74               12.03

          3.87                    Overall 30-day Kit Size     11.62

Note: Kit sizes are in pallets

Figure 9

Airlift Requirement in a Single Deployment

Airlift Requirement in a Single Deployment with:


30-Day Kits                    *****
(66.00 pallets or 3.67 C-17s)

O&ST = Baselines               ***
(50.29 pallets or 2.79 C-17s)

O&ST = 20 Days                 ***
(44.49 pallets or 2.47 C-17s)

O&ST = 15 Days                 **
(32.29 pallets or 1.79 C-17s)

O&ST = 10 Days                 **
(24.02 pallets or 1.33 C-17s)

O&ST = 5 Days                  *
(10.61 pallets or 0.59 C-17s)

Note: 1 C-17 = 18 pallet positions (HQ USAF 2001)

Figure 10

Overall Kit Sizes

Overall Kit Sizes for MRSPs with:


30-Day Kits                      ***********
(573.00 pallets or 31.83 C-17s)  ***********
                                 **********

O&ST = Baselines                 ***********
(418.68 pallets or 23.26 C-17s)  ***********
                                 **

O&ST = 20 Days                   ***********
(362.83 pallets or 20.16 C-17s)  **********

O&ST = 15 Days                   ***********
(267.83 pallets or 14.88 C-17s)  ****

O&ST = 10 Days                   ***********
(194.65 pallets or 10.81 C-17s)

O&ST = 5 Days                    *****
(84.70 pallets or 4.71 C-17s)

Note: 1 C-17 = 18 pallet positions (HQ USAF 2001)

Figure 11

Kit Cost and Savings for a Single Deployment

Kit Cost and Savings for Single Development with:

Figure 11. Kit Cost and Savings for a Single Deployment

Cost with 30-Day Kits    *
($85.51M or 0.36 C-17s)

Saving with              *
O&ST = Baselines
($24.23M or 0.10 C-17s)

Savings with             *
O&ST = 20 Days
($31.20M or 0.13 C-17s)

Savings with             *
O&ST = 15 Days
($45.29M or 0.19 C-17s)

Savings with             *
O&ST = 10 Days
($59.24M or 0.14 C-17s)

Savings with             *
O&ST = 5 Days
($74.60M or 0.32 C-17s)

Note: 1 C-17 = S236.7M (FY98 constant $) (HQ USAF 2001)

Figure 12

Overall Kit Cost and Savings

Overall Kit Cost and Savings for MRSPs with:

Figure 12. Overall Kit Cost and Savings

Cost with 30-Day Kits     ****
($714.86M or 3.02 C-17s)

Saving with               *
O&ST = Baselines
($202.70M or 0.86 C-17s)

Savings with              **
O&ST = 20 Days
($286.26M or 1.13 C-17s)

Savings with              **
O&ST = 15 Days
($391.19M or 1.65 C-17s)

Savings with              ***
O&ST = 10 Days
($506.95M or 2.14 C-17s)

Savings with              ***
O&ST = 5 Days
($628.72M or 2.66 C-17s)

Note: 1 C-17 = S236.7M (FY98 constant $) (HQ USAF 2001)

Notes

(1.) Michael E. Ryan and F. Whitten Peters, America's Air Force: Vision 2020, HQ Air Force, Washington DC, Jun 00, http://www.af.mil/vision/, 20 Aug 00.

(2.) Ibid.

(3.) David Simchi-Levi, Phillip Kaminsky, and Edith Simchi-Levi, Designing and Managing the Supply Chain: Concepts, Strategies, and Case Studies, Boston, Massachusetts: The McGraw-Hill Companies, 2000, 59.

(4.) AFMC Logistics Support Office Logistics Response Time, Wright-Patterson AFB, Ohio [Online] Available: https://137.245.226.104/LRT/database2.htm, 13 Sep 00.

(5.) James T. McClave, P. George Benson, and Terry Sincich, Statistics for Business and Economics, Upper Saddle River, New Jersey: Prentice Hall, Inc. 1998, 55.

(6.) McClave, et al, 58.

(7.) Ibid.

(8.) Michael F. Slay, et al, Optimizing Spares Support: The Aircraft Sustainability Model, AF501MR1, The Logistics Management Institute, McLean, Virginia, Oct 96, 1-1-1-2.

(9.) Slay, et al, 1-3.

(10.) Slay, et al, 1-4.

(11.) Ibid.

(12.) Headquarters, Air Force, "USAF Fact Sheet: C-17 Globemaster III" [Online] Available: http://www.af.mil/new/factsheets/C_17_Globemaster_III.html, 25 Jan 01.

(13.) Correspondence with Capt Eve Burke, Air Combat Command, Weapon System Assessments and Analysis Section, Langley AFB, Virginia, 12-20 Dee 00, and with Capt Daniel Lockhart, Air Mobility Command, Combat Aircraft Support Section, Scott AFB, Illinois, 11-20 Dec 00.

(14.) Jamie D. Allen and Brian Bedesem, "Deploying and Sustaining an F-117A Expeditionary Fighter Squadron: Why Agile Combat Support Is Needed Now," Air Force Journal of Logistics, Vol XXII, No 4, Oct 99.

Major Martinez is chief Wartime Supply and Fuels Analysis Branch, Supply Division, Air Force Logistics Management Agency, Maxwell AFB, Gunter Annex, Alabama. This article is an excerpt from his thesis while a student in the Graduate Logistics Management Program at the Air Force Institute of Technology, Wright-Patterson AFB, Ohio. Lieutenant Colonel Brady is assistant professor of Logistics Management, Air Force Institute of Technology. Major Arostegui is commander, 374th Logistics Readiness Squadron, Yokota Air Base, Japan.

COPYRIGHT 2002 U.S. Air Force, Logistics Management Agency
COPYRIGHT 2004 Gale Group