for Maintenance and Sustainment - DAU 4 - Evaluating Business Models...Business model assumptions. Seven assumptions for the AM business model to follow: • The analysis is from a

The Department of Defense (DoD) must sustain aircraft through the platform’s life cycle. Additive manufacturing could enable DoD organizations to produce spare parts themselves (i.e., organic production), at the point-of-need, anywhere around the globe. For the government to produce spare parts in-house, new business models between the manufacturer and the government have to be established for transacting technical data packages instead of physical spare parts. Can a digital business model be profitable, and can a return on investment (ROI) be achieved within a commercial time horizon? The research includes a survey within government and industry and an analysis using an aviation case study to examine the profitability of four business models. The results show that under the right conditions of nonrecurring costs, digital service costs, and file demand quantities, digital business models can indeed be profitable, although conditions leading to an ROI within 2 or 4 years are limited.

DOI: https://doi.org/10.22594/dau.18-815.26.04 Keywords: 3D Printing, Local Manufacturing, Spare Parts, Technical Data Package, Additive Manufacturing

Image designed by Michael Krukowski

Evaluating Business ModelsENABLING ORGANIC

ADDITIVE MANUFACTURINGfor Maintenance and Sustainment

Ashley N. Totin and Brett P. Connor

382 383Defense ARJ, October 2019, Vol. 26 No. 4 : 380-417 Defense ARJ, October 2019, Vol. 26 No. 4 : 380-417

Evaluating Business Models Enabling Organic Additive Manufacturing https://www.dau.edu October 2019

The Depa r tment of Defense (DoD) faces multiple cha llenges related to obtaining spare parts for weapon systems. One solution can be the manufacturing of spare parts by DoD organizations (i.e., organic manufacturing) using additive manufacturing, which can be performed at the point-of-need and on demand. The goal of this research was to develop and evaluate business models that a manufacturer can provide the government, along with a technical data package (TDP) so that the government can produce spare parts in-house.

The business models, contracting approaches, supply chains, and logistics currently used by the DoD for spare parts are based on conventional manufacturing, which requires items such as tooling and fixturing before

production can start. These items are expensive and have long lead times. This can result in long

lead times for the spare parts themselves (even as long as 800 days for constant

s p e e d d r ive hou si n g c a s t i n g s) ( Mu l l a n , 2 014), s u b s t a nt i a l

a mou nt s of i nventor y ( U. S. Government Accountability

Of f ice [GAO], 2015), long sea rches for suppliers on

legacy systems (Peltz et al., 2014), and critical spare part inventory stock out (Wong, Cattrysse, & Van Oudheusden, 2005)—all of which lead to greater costs and/or reduced mission availability. An emerging

technology called additive manufacturing (AM) enables the reduction of tooling and lead time, which could provide a new avenue for the production of spare parts.

AM is a process in which parts are built by depositing or fusing material, often layer-by-layer (ISO/ASTM, 2015). AM could include nonrecurring costs (NRC) such as design, modeling, or engineering time, and material and machine costs, but AM does not require tooling to start production. As a result, lead-times can be reduced and low-quantity production can become more affordable (Conner et al., 2014). Layer-by-layer fabrication permits complex part designs either impossible or too costly to produce by other means. Lightweighting can then be achieved through topology optimization or lattices and can improve life-cycle costs as material is added only where required (Conner et al., 2014). Customization is also feasible, leading to parts tailored for a mission or operational theater, or parts able to be worn by a specific warfighter (Conner et al., 2014). Using AM, spare parts inventories can be reduced with nearly on-demand part fabrication (Sirichakwal & Conner, 2016).

Project ObjectivesThe goal of this research was to develop and evaluate business models

that a manufacturer can implement when the government needs to produce spare parts in-house. To address the research needs, several research questions were considered:

• How does the profitability of various digital TDP business models compare to the more traditional business model that manufacturers and DoD use? How do the digital TDP models compare to one another? Can manufacturers recoup nonrecurring costs?

• For each business model, how does variability in demand a ffect return on investment (ROI) a nd long-term profitability?

• Does the manufacturing process (conventional versus additive) a manufacturer uses to produce spare parts affect the tota l prof it and years needed to recoup nonrecurring costs?

An emerging technology called additive manufacturing (AM) enables the reduction of tooling and lead

time, which could provide a new avenue for the production of spare parts.



Methodology and ProcedureSurvey

A survey was distributed to personnel from both industry and government to better understand their expectations and the assumptions they make about spare parts and the associated TDP. The survey guided the development of the business models and what types of costs to include. The survey was conducted using the website Survey Monkey. Two sets of questions were established—one for industry members and another for government members. The survey was sent via email to two groups chosen because they were (a) familiar with additive manufacturing and did not require education on the process, (b) were manufacturers or government employees and not solely academics, and (c) were familiar with government contracting and sustainment. The professional groups were the America Makes Maintenance and Sustainment Advisory Group (sent to the aerospace and defense members only) and the Department of Defense Additive Manufacturing Maintenance Operations Group.

The government survey consisted of five questions and the industry survey consisted of five mirroring questions (Table 1). The industry survey included two additional questions because the AM business model research being conducted is to determine how to incentivize industry (Questions 6 and 7).

TABLE 1. SURVEY QUESTIONS & RESULTS

Question 1 Choices Government Industry

Assume a supplier invested in the nonrecurring costs during part development (i.e., design, tooling, qualification, etc.), when do you expect the supplier to obtain a return on investment (ROI)?

< 2 Years 21% 60%

2-4 Years 46% 35%

4-6 Years 29% 5%

6-8 Years 4% 0%

>8 Years 0% 0%

Question 2

Does selling price of a spare part from a supplier change depending on the quantity purchased, or is it a fixed price?

Yes 89% 95%

No 11% 5%

Question 3 0.55 0.77 0.86

If you produced an additive manufactured part in-house, who would pay for the qualification/certification costs?

Government 86% 90%

Industry 14% 10%

Question 4

Consider the situation where the government decided to produce spare parts organically (i.e., at a depot) using additive manufacturing instead of purchasing the spare part. If a supplier provided build files for government to additively produce parts in-house, what type of services would you expect to be included in the selling price? Check all that apply.

24/7 Helpline 25% 25%

Digital Rights Management 68% 75%

Configuration Management 39% 60%

Redesign Services 32% 55%

Update File 61% 60%

Build Process Updates 54% 15%

Secure Storage 43% 65%

Secure Transmission 89% 65%

Field Service Representative 29% 35%

Question 5

How much would you expect to pay for a single copy of a digital technical data package (TDP) file that would allow the government to print a single part?

$0-$500 7% 0%

$500-$2,500 14% 5%

$2,500-$5,000 14% 5%

Question 6

How is profit determined in spare parts production?

Return on Investment (ROI) N/a 10%

Profit Margin added to cost N/a 75%

Other (please specify) N/a 15%

Question 7

How many employees does your company have?

Under 50 N/a 40%

50-100 N/a 5%

200-300 N/a 10%

300-500 N/a 5%

Over 500 N/a 40%



AM Business ModelThe scenario being used here involves a manufacturer that has designed

and currently produces a part, but the government itself wants to produce the same part as a spare part. Although a specific motive is not germane to this scenario, the government may desire this because it wants to produce this part in various global operational theaters, often in remote locations to improve mission availability. For the government to produce this part organically, the digital TDP becomes the transacted item the manufacturer would provide to the government. The manufacturer must now determine a way to maximize value in a situation where it might not be producing the actual part.

Business model assumptions. Seven assumptions for the AM business model to follow:

• The analysis is from a manufacturer’s perspective. The intent is to incentivize the manufacturer to provide data allowing the government to organically produce spare parts. A business model for the government is outside the scope of this research.

• The manufacturer planned to produce spare parts, and therefore invested in tooling, process development, and/or a design to produce the part. The government did not invest in the tooling or development of the part. The manufacturer had planned to recoup the nonrecurring costs (desig n, tooling, etc.) t hroug h spa re pa r ts fabrication.

• The business models assume 1-year prior production of spare parts produced by the manufacturer. In other words, production is underway.

• Models are based on a 15-year forecasted period.

• Industry and government agree upon an established “list price” for files or parts based on estimated demand. Once that list price is set, it does not change.

• Two initia l conditions will be considered: (a) the manufacturer is using conventiona l (nonadditive) manufacturing methods to produce the part, and (b) the manufacturer is using AM methods to produce the part.

• In the case of the conventional manufacturing initial condition, the government will incur the costs of AM qualification and certification; this is a result of the survey that will be addressed later in the article. Those costs and any other government costs are not considered here.

Business model variables. To evaluate the business models, cost models were developed with the variables defined in Table 2. The variables will be used to calculate profit over 15 years for each business model.

TABLE 2. BUSINESS MODEL VARIABLES

Cost Variable Unit Comments

Number of parts produced per year (-) N

Nonrecurring Cost ($) NRC Design/Tooling

Digital Services Cost ($/file) DDigital Thread/24-7 Help Line/Digital Rights Management

Profit Margin (%) PM Based on Aerospace & Defense Profit Margin

Part Cost ($/part) P Cost to Manufacture Part

Carrying Cost ($/year) C Based on 30% value of tooling/or material

Shipping Cost ($/year) SBased on calculations (FedEx, 2016), (FedEx, 2015)(FedEx, 2015, 2016)

Discount Rate (%) DR Based on Aerospace and Defense Market

Period (year) Y Assuming a 15-year period

A 10% profit margin will be the basis of this research. As this is on the high end of acceptable profit margins under the Federal Acquisition Regulation (FAR) (Kinds of Contracts, 1956), a lower margin of 3% will also be examined. Two recurring cost equations were developed using the variables from Table 2: one for where the manufacturer produces the spare parts (either conventionally or additively) and ships them to government (Cmfgr); and one for the scenario where the manufacturer is selling a digital file (Corganic) to the government. These equations will be used to calculate revenue and profit for each model.



Cmfgr = (P*N) + C + S (1)

Corganic = D*N (2)

Model 1: Baseline Manufacturing (BCM or BAM). Model 1 represents the original manufacturing process the manufacturer was using to produce the spare parts. Two different baseline scenarios will be analyzed: (a) first a baseline conventional manufacturing process, which represents a traditional manufacturing process to produce spare parts; and (b) a baseline AM model, which represents a scenario where the manufacturer was already producing the spare parts utilizing AM. Both baselines will be used to compare data obtained from business models 2–4. Tooling costs associated with the BCM are amortized within the first 2 years. It is assumed that the tooling needed for the manufacturing process is a one-time purchase for the part’s entire life cycle, due to low-volume production.

Revenue (RBM) and profit (PBM) equations for BCM and BAM are shown below.

RBM = Cmfgr + (Cmfgr * PM) (3)

PBM = RBM - Cmfgr (4)

Model 2: One Time Sale (OTS). OTS gives the government the opportunity to buy the TDP outright. The government owns all responsibility and rights to the part. Conversely, the manufacturer is relieved of all responsibility and rights, but also relieved of actual earned profit from spare parts sales. OTS profit is calculated using the net present value (NPV) formula. The manufacturer can use NPV to determine the value of selling the TDP outright by determining the future profits they will attain. The NRC is amortized within the first 2 years of the 15-year period used in the NPV. The NRC will be subtracted from the NPV total profit values shown in the results. In equation (5) (Ostwald and McLaren, 2004), profit per year is based on an assumption of the quantity of spare parts sold. The discount rate (DR) utilized in this study is 8%, based on the aerospace and defense market.

POTS = NPV = ∑ - NR (5)

Model 3: File Per Use (FPU). FPU allows for the manufacturer to continue owning the rights of the spare part. When the government wants to produce a part on demand using AM, they will purchase a TDP file from the manufacturer. The file will be accessible for one build, and then it is no longer available. If the government wants to produce a second part, another TDP file will be purchased. Included in the cost per file are digital services, which can include customer support, data storage, cybersecurity, blockchain, and TDP updates. The TDP is expected to evolve over the life cycle of the weapons systems and will need to be updated. Some potential causes of change include: a new operating system for AM machine, materials supplier change, AM machine obsolescence, and a design change to the part. Revenue (RFPU) and profit (PFPU) equations for FPU are shown below.

RFPU = Corganic + (Corganic + PM) (6)

PFPU = RFPU - Corganic (7)

Model 4: Annual Subscription Fee (ASF). ASF is similar to FPU as it is a business model based on TDP file transactions. ASF, however, is based on an annual subscription fee. The manufacturer will charge an annual subscription fee and the government can download an unlimited number of TDP files. The annual fee (AF) is determined by calculating the average quantity of files (Nestimated) the government will request per year and multiplying the number by the digital services cost per file (D), illustrated in equation 8.

AF = Nestimated *D (8)

Yy=1 Profity

(1+DR)y



Revenue and profit equations for ASF are shown below.

RASF = AF + (AF*PM) (9)

PASF = RASF - Corganic (10)

It should be noted from equation 10 that although revenue is given in one annual lump sum, the recurring costs are realized on a per file basis.

Case study selection. A yoke cover (consisting of two halves: right and left) from a legacy airlift platform was selected for the case study (Figure 1).

FIGURE 1. YOKE COVER CONSISTING OF TWO HALVES

Note. Photo courtesy Youngstown State University.

This part was selected because of the extensive lead time associated with obtaining spares. The yoke cover was reverse engineered by 3D scanning with a 3D Systems Capture scanner to obtain geometric point cloud data. After scanning, Geomagic Design X software was used to convert the point cloud data into a digital solid model suitable for 3D printing. The design time was 38 hours. The design time would be included in the NRC for the AM models.

The yoke cover was conventionally manufactured by injection molding (IM), which involves injecting a molten polymer into a mold, where the part solidifies and then the mold opens. Although the cycle time per part is very short (seconds or at most minutes), the lead time for tooling can be months and the cost of tooling can be tens of thousands of dollars. The cost to produce the yoke cover with IM was calculated using an IM cost estimator from the website Custompart.Net (2019), and is summarized in Table 3. The costs shown in Table 3 are for one half of the yoke cover; therefore, the costs will be doubled to account for the entire part.

TABLE 3. INJECTION MOLDING COSTS FOR YOKE COVER

Production Volume (pcs) N

Envelope X-Y-Z in 3.14 x 2.95 x 0.59

Material cost per part $ 0.25

Production cost per part $ 2.4

Tooling cost $ 30,000/N

Note. (CustomPart.net, 2019)

For the AM approach, a Stratasys Fortus 900mc material extrusion AM printer was chosen, and the material selected was ULTEM 9085 thermoplastic. To maximize the capacity of the machine, the build is packed with 20 yoke covers (each with right and left parts) oriented to minimize support structures. The build time is calculated using the Stratasys Insight and Control Center software. The parameters and costs associated with the Fortus 900mc and the production of the yoke cover are summarized in Table 4.



TABLE 4. FORTUS 900MC COSTS FOR YOKE COVER

Number of Parts N 20

Machine Consumable Parameters

Print Material used per build mm3 1,638,000

Support Material used per part build mm3 398,000

Machine Parameters

X Printable Dimension mm 914.4

Y Printable Dimension mm 609.6

Z Printable Dimension mm 914.4

Layer Thickness µm 254.0

Build Time hr. 94.1

Preprocessing hr. 1.0

Postprocessing hr. 2.0

Costs

Print Material $/cc 0.45

Support Material $/cc 0.45

Support Removal Cost $ 25.00

Machine Cost $/hr. 10.37

Machine Operator Cost $/hr. 25.00

Build Cost

Print Material Cost $ 742.94

Support Material Cost $ 205.49

Preprocessing Cost $ 25.00

Postprocessing Cost $ 75.00

Machine Time Cost $ 976.57

Recurring Cost/Build $ 2,025.00

Recurring Cost/Part $ 100.00

An Air Force Reserve airlift wing provided an average annual demand of 2–5 yoke cover sets for eight aircraft. A mean demand was calculated as 3.5 yoke covers per eight aircraft (0.4375 yoke covers/aircraft). Applied to the total fleet of 290 aircraft, this results in an annual mean demand of approximately 130 yoke cover sets for the entire fleet. A Monte Carlo simulation was used to consider the effects of uncertainty on demand (Winston, 2016). Standard deviations were calculated for both the number of yoke covers and the number of aircraft in a fleet. These standard deviations were then used to calculate an estimated demand for 15 years. Two other demand rates were considered with calculated standard deviations. The following demand rates will be used to calculate profit: 130 Fixed, 130±38, 92 ± 38, 168 ± 38. From the results section, individuals agreed the cost per TDP file will vary; therefore, for this research, the Digital Service Cost will be varied from $250, $500, and $1,000. The cost elements that will be used in this research are shown in Table 5.

TABLE 5. COST ELEMENTS

Models 1-4

Nonrecurring Costs $30,000.00

Tooling Cost $60,000.00

BCM & BAM

Carrying Cost

Injection Molding $18,000.00

Material Extrusion $2,055.00

Shipping Cost

Injection Molding $600.00

Material Extrusion $160.00

Production Cost

Injection Molding $5.20

Material Extrusion $100.00

FPU & ASF

Digital Service Cost/File $250, $500, & $1,000



Analysis and ResultsSurvey Results

The business model survey had a total of 48 responses—28 from govern-ment members and 20 from industry members. Forty percent of industry respondents came from businesses of less than 50 people and 40% came from companies larger than 500 people. The results of the survey can be seen in Table 1; additional comments and discussion can be found in the appendix.

An important conclusion from the survey is 75% of industry members agree profit is obtained through a profit margin. Therefore, this research will calculate profit based on a profit margin. The results of the survey also conclude that a 2-year ROI is expected by the manufacturer; therefore, an analysis on what profit margin is needed to obtain a 2-year ROI will be conducted. A strong majority of participants (both industry and government) stated that selling price is a function of quantity. This differs from our preceding model assumption, but the assumption is a reasonable approximation in that it is more in line with the fact that AM fabrication costs are far less sensitive to quantity than conventional manufacturing (Almaghariz et al., 2016; Atzeni & Salmi, 2012; Hopkinson & Dickens, 2003; Ruffo & Hague, 2007). The survey shows that both government and industry participants understand that “one-size-fits-all” TDP prices are not realistic and prices will reflect the nature (i.e., complexity, criticality, etc.) of the actual part.

Business Model ResultsThe objective of this study was to calculate the total profit over 15 years

and the time needed to recoup the NRC for all the business models and compare the results. The business models will be analyzed through two scenarios: (a) industry originally produced spare parts with conventional manufacturing, and (b) industry-produced spare parts with AM.

Initial condition: Manufacturer uses conventional manufacturing. The four models are now evaluated where the manufacturer is using conven-tional manufacturing (BCM). In this case, injection molding is used, which requires tooling and therefore a large NRC ($90,000). Selling prices were calculated using equations 3, 6, and 9 for BCM, FPU, and ASF, respectively. Once the selling prices per part or per file are calculated for the specified quantity of 130, the price becomes fixed as noted in the assumptions. The fixed 130 represents the forecasted demand the manufacturer assumes when calculating the selling price. The other three demand variations represent the actual quantity of parts requested from the government each year for 15 years. First, we will assume that the revenue for the profit margin model is obtained by adding the costs to produce the part or file plus an additional margin—both 10% and 3% margins (Tables 6 and 7, respectively) were ana-lyzed. The two tables display the total profit and time to recoup NRC for BCM, FPU, and ASF. Four different demand variations are utilized. The profit totals in the tables do not include subtracting out the NRC.

For OTS, the NPV calculation yields a profit of $89,131 for 10% margin and $19,122 for 3% margin—both totals include subtracting out the NRC. OTS can exceed the profit from BCM in the case of lower than anticipated parts demand. In the presence of demand variation, FPU is often less lucrative than BCM or OTS. ASF can result in substantial loss in the presence of higher than anticipated file demand (with correspondent associated per file services costs) or simply unanticipated costs (i.e., unexpected file changes, etc.). The manufacturer is unlikely to achieve an ROI within a 2- or 4-year window in nearly all of the cases involving digital models and, for most cases, would not recoup the initial investments within 5 years. To find the profit margin necessary to achieve a 2-year ROI, a simulation was run using the Goal Seek application in Excel software to cycle through all possible profit margins until the profit for year 2 returns the specified nonrecurring cost. As shown in Table 8, the profit margins needed to achieve an ROI within 2 years are well above those allowed for FAR contracts (except for the ASF model where, in a few cases, demand turned out to be lower than anticipated). Models that achieve an ROI in 2 years or less are represented by NC, which means no change.



TABLE 6. FORECASTED DEMAND VS. ACTUAL DEMAND (INITIAL TRADITIONAL) PROFIT MARGIN 10%

Total Profit 15 Years Total Profit 15 Years

Demand BCM

FPU ASF

Digital Services $250/file


Digital Services $1,000/file

Digital Services$250/file


Digital Services$1,000/file

Fixed 130 $169,914 $48,750 $97,500 $195,000 $48,750 $97,500 $195,000

130 ± 38 $220,734 $48,400 $96,800 $193,600 $52,250 $104,500 $209,000

92 ± 38 $79,229 $32,825 $65,650 $131,300 $208,000 $416,000 $832,000

168 ± 38 $376,776 $65,575 $131,150 $262,300 -$119,500 -$239,000 -$478,000

Total Years to Recoup NRC ($90,000)

Demand BCM

FPU ASF







Fixed 130 3 >15 14 7 >15 14 7

130 ± 38 3 >15 15 8 >15 12 4

92 ± 38 8 >15 >15 10 9 5 1

168 ± 38 2 >15 11 6 >15 >15 >15

TABLE 7. FORECASTED DEMAND VS ACTUAL DEMAND (INITIAL CONVENTIONAL) PROFIT MARGIN 3%


Demand BCM

FPU ASF







Fixed 130 $50,974 $14,625 $29,250 $58,500 $14,625 $29,250 $58,500

130 ± 38 $106,818 $14,520 $29,040 $58,080 $18,125 $36,250 $72,500

92 ± 38 $1,971 $9,848 $19,695 $39,390 $173,875 $347,750 $695,500

168 ± 38 $222,436 $19,673 $39,345 $78,690 -$153,625 -$153,625 -$307,250 -$614,500


Demand BCM

FPU ASF







Fixed 130 9 >15 >15 >15 >15 >15 >15

130 ± 38 8 >15 >15 >15 >15 >15 13

92 ± 38 >15 >15 >15 >15 10 6 2

168 ± 38 3 >15 >15 >15 >15 >15 >15



Total profit ratio (profit/NPV) as a function of mean quantity was used to compare OTS profit with the other options’ total profit. Mean quantities from 92 to 168 were selected, and a normal random distribution over 15 years was applied for a total of 76 total profits. Nonrecurring costs were subtracted from the total profit to be consistent with the NPV equation. These profits were then divided by the NPV. A value greater than one means the option obtains a greater profit than OTS, and a value lower than one signifies the total profit is less than OTS. A value is achieved, which represents the ratio of the total profit/NPV on the y-axis and the mean quantity of data (or parts in the case of BCM on the x-axis). The results can be seen in Figures 2 and 3.

FIGURE 2. PROFIT/NPV—INITIAL CONVENTIONAL (10% MARGIN)

-6

-4

-2

0

2

4

6

92 95 98 101

104

107

110

113

116

119

122

125

128

131

134

137

140

143

146

149

152

155

158

161

164

167

PRO

FIT/

NPV

MEAN QUANTITY

Initial Condition: Conventional Manufacturing Profit Margin: 10%

OTS

BCM (St. Dev. ± 0.93)

FPU: $250 (St. Dev. ± 0.10)

FPU: $500 (St. Dev. ± 0.20)

FPU: $1,000 (St. Dev. ± .41)

ASF: $250 (St. Dev. ± 1.02)

ASF: $500 (St. Dev. ± 2.04)

ASF: $1,000 (St. Dev. ± 4.08)

FIGURE 3: PROFIT/NPV—INITIAL CONVENTIONAL (3% MARGIN)

-10-8-6-4-202468

10

92 97 102

107

112

117

122

127

132

137

142

147

152

157

162

167

PRO

FIT/

NPV

MEAN QUANTITY

Initial Condition: Conventional Manufacturing Profit Margin: 3%OTS

BCM (St. Dev. ±3.07)

FPU: $250 (St. Dev. ± 0.14)

FPU: $500 (St. Dev. ± 0.27)

FPU: $1,000 (St. Dev. ± 0.55)

ASF: $250 (St. Dev. ± 4.57)

ASF: $500 (St. Dev. ± 9.13)

ASF: $1,000 (St. Dev. ± 18.26)

Since profit is added to costs, when the digital service costs are lower (say $250 per file), FPU and ASF models do not return a higher profit than sell-ing the data outright (OTS). The FPU model performs better as demand increases, but only outperforms OTS when the digital services costs are at $1,000 per file. Additionally, ASF only is profitable when demand turns out to be lower than expected (and digital services costs are $500 or 1,000 per file). Thus, the ASF model should not be implemented if a high-demand variation is forecasted, as the manufacturer would risk losing money.

If the profit margin is lowered to 3%, then the ROI is pushed out beyond 15 years for BCM and FPU at all quantities, and only a few conditions of ASF have an ROI less than 15 years in the case of lower than anticipated demand.

Initial condition: Manufacturer uses AM. Now the initial condition is changed where the manufacturer has been producing the yoke covers using FDM AM instead of conventional injection molding. The BAM scenario for Model 1 will be utilized. There are still NRCs related to design and process setup, but without tooling the NRC is only 1/3 of that for the conventional manufacturing case ($30,000). For the OTS model, the new NPV calculation yields a profit of $35,828 (10%) and $8,472 (3%). Tables 9 and 10 show total profit over 15 years for all four models and also the years needed to recoup ROI.

TABLE 8. PROFIT MARGIN FOR 2-YEAR ROI—INITIAL CONVENTIONAL

Demand BCMFPU ASF

$250/file $500/file $1,000/file $250/file $500/file $1,000/file

130 Fixed 14% 138% 69% 35% 138% 69% 35%

130 ± 38 16% 148% 74% 37% 132% 63% 28%

92 ± 38 38% 250% 125% 63% 94% 25% NC

168 ± 38 NC 119% 60% 30% 154% 85% 50%



TABLE 9. FORECASTED DEMAND VS ACTUAL DEMAND (INITIAL ADDITIVE) PROFIT MARGIN: 10%


Demand BAMFPU ASF







Fixed 130 $68,482 $48,750 $97,500 $195,000 $48,750 $97,500 $195,000

130 ± 38 $67,097 $48,400 $96,800 $193,600 $52,250 $104,500 $209,000

92 ± 38 $34,814 $32,825 $65,650 $131,300 $208,000 $416,000 $832,000

168 ± 38 $102,697 $65,575 $131,150 $262,300 -$119,500 -$239,000 -$478,000


Demand BAM

FPU ASF







Fixed 130 7 10 5 3 10 5 3

130 ± 38 7 9 5 3 6 5 2

92 ± 38 12 14 7 4 2 2 1

168 ± 38 5 8 4 2 >15 >15 >15

TABLE 10. FORECASTED DEMAND VS ACTUAL DEMAND (INITIAL ADDITIVE) PROFIT MARGIN: 3%


Demand BAMFPU ASF







Fixed 130 $20,545 $14,625 $29,250 $58,500 $14,625 $29,250 $58,500

130 ± 38 $19,962 $14,520 $29,040 $58,080 $18,125 $36,250 $72,500

92 ± 38 $2,847 $9,848 $19,695 $39,390 $173,875 $347,750 $695,500

168 ± 38 $38,836 $19,673 $39,345 $78,690 -$153,625 -$153,625 -$307,250 -$614,500


Demand BAM

FPU ASF







Fixed 130 >15 >15 >15 8 >15 >15 8

130 ± 38 >15 >15 >15 8 13 6 4

92 ± 38 >15 >15 >15 11 2 2 1

168 ± 38 13 >15 12 6 >15 >15 >15



When starting with AM, the total years to recoup NRC is far shorter through FPU, and ASF is substantially shorter than when starting with conventional manufacturing. In fact, in most cases ROI is achieved between 1 to 5 years. Not only that, but in most cases the digital business models (FPU and ASF), where government procures files instead of parts, achieved ROI faster than the physical BAM business model where government procured parts from the manufacturer. Table 11 is another analysis of what level of profit margin would be needed to obtain an ROI of 2 years or less but for an initial con-dition involving AM. The profit margins in Table 11 are lower than those shown in Table 8, but they are still higher than typical for FAR contracts.

Examining the total profit ratio as a function of mean quantity for each model shows that the FPU model normalized profits are nearly equivalent or better (even up to 7X) compared to OTS and BAM for nearly all quantities regardless (see Figures 4 and 5). However, ASF profitability is very sensitive to mean quantity.

FIGURE 4. PROFIT/NPV—INITIAL ADDITIVE MANUFACTURING (10% MARGIN)

-10

-8

-6

-4

-2

0

2

4

6

8

10

92 96 100

104

108

112

116

120

124

128

132

136

140

144

148

152

156

160

164

168

PRO

FIT/

NPV

MEAN QUANTITY

Initial Condition: Additive Manufacturing Profit Margin: 10%OTS

BAM (St. Dev. ± 0.53)

FPU: $250 (St. Dev. ± 0.25)

FPU: $500 (St. Dev. ± 0.51)

FPU:l $1,000 (St. Dev. ± 1.02)

ASF: $250 (St. Dev. ± 2.54)

ASF: $500 (St. Dev. ± 5.08)

ASF: $1,000 (St. Dev. ± 10.15)

FIGURE 5. PROFIT/NPV—INITIAL ADDITIVE MANUFACTURING (3% MARGIN)

-10-8-6-4-202468

10

92 97 102

107

112

117

122

127

132

137

142

147

152

157

162

167

PRO

FIT/

NPV

MEAN QUANTITY

Initial Condition: Additive Manufacturing Profit Margin: 3%OTS

BCM (St. Dev. ± 1.18)

FPU: $250 (St. Dev. ± 0.32)

FPU: $500 (St. Dev. ± 0.64)

FPU $1,000 (St. Dev. ± 1.29)

ASF: $250 (St. Dev. ± 10.73)

ASF: $500 (St. Dev. ± 21.46)

ASF: $1,000 (St. Dev. ± 42.92)

Discussion and ConclusionsSurvey responses of industry and government members were used to set

a baseline for the business models. In the survey, industry indicated that a 2-year ROI was expected. Government and industry participants agreed on the government paying for certification and qualification costs, which was also assumed. Then an analysis investigated four business models in the

TABLE 11. PROFIT MARGIN FOR 2-YEAR ROI—INITIAL ADDITIVE

Demand BAMFPU ASF

$250/file $500/file $1,000/file $250/file $500/file $1,000/file

130 Fixed 33% 46% 23% 12% 46% 23% 12%

130 ± 38 36% 49% 24% 12% 40% 17% NC

92 ± 38 64% 82% 42% 21% NC NC NC

168 ± 38 28% 40% 20% NC 62% 39% 27%



scenario in which the government wants to use industry technical data to move from industry production of a spare part to government production of spare parts in-house. The digital services cost was varied to account for the varied responses on what should be included in the services offered to the government and the cost of the TDP. Models establishing price based on a fixed profit margin model were utilized due to the survey responses.

Key Findings• T he resu lt s show u nder t he r i g ht cond it ions of

nonrecurring costs, digita l ser vice costs, and f ile demand quantities, digital business models can indeed be profitable although conditions leading to an ROI within 2 or 4 years are limited. All analysis results are a function of total quantity of parts and demand rate.

• Sixty percent of industry participants expected an ROI in less than 2 years, and approximately 80% of government participants selected 4 or more years for an ROI. While this is a disagreement, there is an opportunity for negotiation to attempt to achieve an ROI between 2 and 4 years. As shown in the analysis, achieving an ROI in even a 4-year timeframe might be difficult. The profit margins are well above what FAR allows.

• For an initial condition of conventional manufacturing, digital TDP business models can be as profitable as or more profitable than BCM as long as digital services costs are sufficiently high (since profit is added to cost). The exception being ASF where demand is higher than anticipated, resulting in significant loss.

• In general, obtaining an ROI within the 2 years (even 4 years) identified in the survey is difficult regardless of method. If such a timeframe is necessary, only OTS will guarantee it. The analysis does show that an initial condition where the manufacturer uses AM increases the opportunity for ROI to be achieved within 2 to 4 years. Alternatively, the government could incentivize a digital TDP transaction model by proposing reimbursement for some or all of the NRC.

• For a constant profit margin, it can be shown that compared to the other models, the annual subscription fee (ASF) model as defined here is highly sensitive to

variation in quantity and therefore is more risky. This sensitivity increases when there are: (a) lower profit margins and/or (b) higher digital service costs. The FPU model is less sensitive to variation in quantity. Like the BCM/BAM model, FPU profit increases with more quantity demand and decreases with less. However, FPU is less sensitive to quantity variation than BCM/BAM as the smaller profit/NPV versus quantity slope indicates.

• Interestingly, if the manufacturer can only obtain a low profit margin and if digital services costs are not high, the outright sale of the TDP (OTS) would be the best approach for the manufacturer. The outright sale of the TDP has several advantages in that it minimizes the time to recoup NRC; it reduces production-related liability to the manufacturer (although not necessarily design liability); it eliminates costs associated with a production infrastructure or a digital data infrastructure; and it guarantees profit now in light of future uncertainty. It is shown to be rarely the most profitable method long-term. However, it is possible for the manufacturer to reinvest the funds and improve the long-term profit although this scenario is not shown here.

• An initial condition of AM has a lower NRC than an initial condition of conventional manufacturing; therefore, the NRC can be recouped much faster using FPU and ASF if the profit margin is sufficiently high. Additionally, when the initial manufacturing process is AM, lower digital service costs are needed for FPU (and ASF as long as higher than expected quantities aren’t encountered) to be more profitable than BAM. Interestingly, BCM is more profitable

An implication can be that if a company that has invested in tooling and is in conventional manufacturing, it may be

less willing to allow the government to establish an approach of organic manufacturing. However, a company that starts with AM may be more willing to share production with the government.



than BAM. An implication can be that if a company that has invested in tooling and is in conventional manufacturing, it may be less willing to allow the government to establish an approach of organic manufacturing. However, a company that starts with AM may be more willing to share production with the government.

Further Discussion and Recommendations• For new weapon systems, if gover nment desires

organic A M spare part production, this should be acknowledged early in the acquisition process so that ROI and profitability concerns can be addressed. The implementation of AM into the production plan provides benefits to both the manufacturer and the government.

• O t her Tra nsaction Aut hor it y (OTA) a g reement s (Authority of the Department of Defense, 1989) could be considered for implementing digital TDP models. OTA is more f lexible towards profit and also enables nontraditional participants. For a more traditional approach, Performance Based Logistics (PBL) contracts may leverage the AM capabilities within depots.

• A n O T S m o d e l c o u l d f o s t e r a n o n t r a d i t i o n a l manufacturing partner who has digital design services and AM capabilities, but oriented towards prototyping and not engaging in production itself. Instead, this company would sell the TDP to the government. With conventional manufacturing assets, this would be highly unusual, but digital manufacturing can enable it. However, a potential barrier for some companies to conduct an OTS would be if the government turned around and gave the TDP to another company who could be a competitor of the original designing firm. The government needs to take care not to dis-incentivize this approach.

• In a digital TDP transaction model, as long as the manufacturer retains rights to the TDP, it will require a digita l infrastructure (with associated costs) to securely store and transmit TDP data (even unclassified), verify data validity, control configuration, update files, provide customer support, and facilitate transactions (to include Digital Rights Management, or DRM). However,

the government also needs a digital infrastructure to implement organic AM. There can be an opportunity for cost savings to both parties with shared infrastructure although both entities will have their own infrastructure requirements.

• It should be recognized that although the case study examined yoke covers produced additively using material extrusion, these results are extensible to other additive processes. The authors examined a study on a metal landing gear component produced conventionally using die casting and additively using laser powder bed fusion (Atzeni & Salmi, 2012). Applying demand data for an actual part with substantial variation over a 10-year period found in Peltz et al. (2014), the business model results were the same.

• Areas of future research are numerous. One is an analysis of business models from the government’s standpoint that determine the value to the government. For example, Tables 12 through 15 show the purchase price by the government. W hile government production would have logistical advantages of on-demand, on-location production and the cost advantages of eliminated stockpiles, would the prices shown in Tables 12 through 15 be acceptable to the government? One should expect the government to negotiate the pricing during the process of reaching an agreement.

• Another future research area would be exa mining the una nticipated consequences of DoD production if manufacturer also sells the part to commercial customers or foreign military sales. Additionally, other digital model s shou ld be a na ly zed—hybr id combinations of FPU and APF, or a model where prices slide as a function of quantity.



TABLE 12. SELLING PRICES TO GOVERNMENT (INITIAL TRADITIONAL) PROFIT MARGIN 10%


Demand BAMFPU ASF







Fixed 130 $513,054 $536,250 $1,072,500 $2,145,000 $536,250 $1,072,500 $2,145,000

130 ± 38 $509,801 $532,400 $1,064,800 $2,129,600 $536,250 $1,072,500 $2,145,000

92 ± 38 $365,056 $361,075 $722,150 $1,444,300 $536,250 $1,072,500 $2,145,000

168 ± 38 $669,416 $721,325 $1,442,650 $2,885,300 $536,250 $1,072,500 $2,145,000

TABLE 13. SELLING PRICES TO GOVERNMENT (INITIAL TRADITIONAL) PROFIT MARGIN 3%


Demand BAMFPU ASF







Fixed 130 $398,314 $502,125 $1,004,250 $2,008,500 $502,125 $1,004,250 $2,008,500

130 ± 38 $395,885 $498,520 $997,040 $1,994,080 $502,125 $1,004,250 $2,008,500

92 ± 38 $287,798 $338,098 $676,195 $1,352,390 $502,125 $1,004,250 $2,008,500

168 ± 38 $515,076 $675,423 $1,350,845 $2,701,690 $502,125 $1,004,250 $2,008,500

TABLE 14. SELLING PRICES TO GOVERNMENT (INITIAL ADDITIVE) PROFIT MARGIN 10%


Demand BAMFPU ASF







Fixed 130 $296,048 $536,250 $1,072,500 $2,145,000 $536,250 $1,072,500 $2,145,000

130 ± 38 $293,922 $532,400 $1,064,800 $2,129,600 $536,250 $1,072,500 $2,145,000

92 ± 38 $199,339 $361,075 $722,150 $1,444,300 $536,250 $1,072,500 $2,145,000

168 ± 38 $398,222 $721,325 $1,442,650 $2,885,300 $536,250 $1,072,500 $2,145,000

TABLE 15. SELLING PRICES TO GOVERNMENT (INITIAL ADDITIVE) PROFIT MARGIN 3%


Demand BAMFPU ASF







Fixed 130 $248,572 $502,125 $1,004,250 $2,008,500 $502,125 $1,004,250 $2,008,500

130 ± 38 $246,787 $498,520 $997,040 $1,994,080 $502,125 $1,004,250 $2,008,500

92 ± 38 $167,372 $338,098 $676,195 $1,352,390 $502,125 $1,004,250 $2,008,500

168 ± 38 $334,361 $675,423 $1,350,845 $2,701,690 $502,125 $1,004,250 $2,008,500



References

Mullan, R. (2014). 422nd SCMS provides key support to the warfighter. Tinker Take Off, 72(39). Retrieved from http://journalrecord.com/tinkertakeoff/2014/09/25/422nd-scms-provides-key-support-to-the-warfighter/

Almaghariz, E. S., Conner, B. P., Lenner, L., Gullapalli, R., Manogharan, G. P., Lamoncha, B., & Fang, M. (2016). Quantifying the role of part design complexity in using 3D sand printing for molds and cores. International Journal of Metalcasting, 10(3), 240–252. Retrieved from https://doi.org/10.1007/s40962-016-0027-5

Atzeni, E., & Salmi, A. (2012). Economics of additive manufacturing for end-usable metal parts. International Journal of Advanced Manufacturing Technology, 62(9), 1147–1155. Retrieved from http://www.springerlink.com/index/K08M081535085168.pdf

Authority of the Department of Defense to Carry Out Certain Prototype Projects, 10 U.S.C. § 2371b (1989). Retrieved from https://www.govinfo.gov/app/details/USCODE-2015-title10/USCODE-2015-title10-subtitleA-partIV-chap139-sec2371b

Conner, B. P., Manogharan, G. P., Martof, A. N., Rodomsky, L. M., Rodomsky, C. M., Jordan, D. C., & Limperos, J. W. (2014). Making sense of 3-D printing: Creating a map of additive manufacturing products and services. Additive Manufacturing, 1–4, 64–76. Retrieved from https://doi.org/10.1016/j.addma.2014.08.005

Custompart.Net. (2019). Injection molding cost estimator. Retrieved from http://www.custompartnet.com/estimate/injection-molding/

Hopkinson, N., & Dickens, P. (2003). Analysis of rapid manufacturing—using layer manufacturing processes for production. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 217(1), 31–39. Retrieved from https://doi.org/10.1243/095440603762554596

ISO/ASTM. (2015). In Standard terminology for additive manufacturing—general principles – terminology, 1st ed. (No. 52900-15). West Conshohocken, PA: ASTM International.

Kinds of Contracts, 10 U.S.C. § 2306 (1956).Ostwald, P. F., & McLaren, T. S. (2004). Cost analysis and estimating for engineering and

management. Upper Saddle River, NJ: Pearson/Prentice Hall. Retrieved from https://market.android.com/details?id=book-xYZRAAAAMAAJ

Peltz, E., Brauner, M. K., Keating, E. G., Saltzman, E., Tremblay, D., & Boren, P. (2014). DoD depot-level reparable supply chain management: Process effectiveness and opportunities for improvement. Retrieved from http://www.dtic.mil/docs/citations/ADA602842

Ruffo, M., & Hague, R. (2007). Cost estimation for rapid manufacturing ’ simultaneous production of mixed components using laser sintering. Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture, 221(11), 1585–1591. Retrieved from https://doi.org/10.1243/09544054JEM894

Sirichakwal, I., & Conner, B. (2016). Implications of additive manufacturing for spare parts inventory. 3D Printing and Additive Manufacturing, 3(1), 56–63. Retrieved from https://doi.org/10.1089/3dp.2015.0035

U.S. Government Accountability Office. (2015). Defense inventory: Services generally have reduced excess inventory, but additional actions are needed (Report No. GAO-15-350). Washington, DC: Author.

Winston, W. (2016). Microsoft Excel data analysis and business modeling. Microsoft Press. Retrieved from https://market.android.com/details?id=book-ZbmPDQAAQBAJ

Wong, H., Cattrysse, D., & Van Oudheusden, D. (2005). Stocking decisions for repairable spare parts pooling in a multi-hub system. International Journal of Production Economics, 93-94, 309–317. Retrieved from https://doi.org/10.1016/j.ijpe.2004.06.029

AppendixSurvey Results: Open Response

Question 1: Assume a supplier invested in the nonrecurring costs during part development (i.e., design, tooling, qualification, etc.), when do you expect the supplier to obtain a return on investment (ROI)?

Additional comments by government:• “This will vary depending on size of investment and BCA [business case

analysis] results.”• “[Two years or less] is standard for stock-held companies or those owned by

a hedge fund.”• “Depends on the life of the weapon system the part is associated with.”• “Military budgets in 5-year cycles called FYDP [Future Years Defense Plan];

therefore, I would expect [an] ROI within the FYDP.”• “[My] assumption is that an economical buy was executed for the effort.”• “Longer than that is too risky for a company unless they are getting a steady

stream of external funding and investment.”

Additional comments from industry:• “It would be included with the first part unless there is a contracted number

to buy.”• “It really depends on the nature and amount of the expense to give a better

answer, but I think 18 to 14 months is enough time to see a return.”• “Depends on platform and quantity. If there are international customers, the

ROI would be quicker.”

Question 1 discussion: From the graph (Table 1) and the comments, the conclusion is a timeframe of more than 4 years is not acceptable for a return on investment from an industry standpoint. There does seem to be some difference between respondents from government as compared to those from industry on how quickly ROI should be achieved. While 60% of industry participants expect less than 2 years’ ROI, only 20% of the government participants expect industry to achieve an ROI in less than 2 years. Forty-six percent of government participants expect a 2- to 4-year ROI. The comment shown from a government participant suggested that an ROI within the timeframe of the FYDP was sufficient—despite uncertainties in quantities that can occur during a FYDP, including system cost increases or congressional reduction in funding. This difference in expectation could complicate a negotiation for data rights between the government and industry. The majority of industry participants expect to see an ROI in less than 2 years; therefore, a 2-year ROI is selected for the business model analysis.

Question 2: Does selling price of a spare part from a supplier change depending on the quantity purchased, or is it a fixed price?



Additional comments from government:• “This again can be an either or ... however, in general, buying bulk of traditional

manufactured processes items usually lends itself to cost reduction. AM is unknown.”

• “This depends on the spare part and how it is manufactured. Most conventional manufacturing methods would typically exhibit volume ordering discounts.”

• “Selling price should be fixed by some outcome-based product support strategy (i.e., Performance Based Logistics). AM requires less tooling/set-up for low demand orders; small orders can still be profitable for supplier.”

• “Nonrecurring costs should be spread over the initial quantity acquired.”• “With AM, it wouldn’t change nearly as much, but nevertheless, there should

still be definitive quantity price breaks regardless.”• “Yes, since volume defines the utilization rate of the processes. It drives

business decisions.”

Additional comments from industry:• “Amortized the nonrecurring costs plus transaction costs.”• “Unless it is a reorder and no NRE [nonrecurring engineering] is anticipated.”• “Selling price is fixed on a projected quantity. Some quantity would need to be

contractual.”• “Out of production, yes. Production no.”

Question 2 discussion: The overall consensus is quantity does determine the selling price of a product. Many of the comments conclude price would be based on a specified volume. The results of this question helped determine how variation in demand would affect the business models. Based on these answers, the business models developed used a specified estimated demand (130 Yoke Covers) the government expected to need per year. The demand was used to calculate the selling price per part. Once the price was calculated, the price becomes a fixed firm price and does not change. Now if the demand changes per year, having a fixed firm price will allow for an analysis of the different models and how a demand variation affects the models.

Question 3:If you produced an additive manufactured part in-house, who would pay for the qualification/certification costs?

Additional comments from government:• “Ultimately the cost is passed to the government, either as a direct cost or as a

built-in cost masked as something else.”• “If the customer wants a service, government or not, you pay for that service.”• “In the long term, the government will not be in the business of paying for

the qualification & certification of AM parts. However, the current state of the industry is such that widely publicized, statistically significant AM material properties (metal or polymer) plus the relevant industry standards are not available. Therefore, there are and will be instances where industry and government partner to achieve qualification & certification of a part, whether that be a point solution or a process solution.”

• “Depends on complexity, critical risk to life and system, and who owns the technical requirements the part complies with.”

• “Government should pay for what they use, especially if they produce an AM part.”

• “Industry would provide inspection and acceptance points for government test & acceptance. FAR 52.209-4 & 9.308-2”

• “It all depends on who is doing the work and where the requirement comes from. If industry looks at AM to supplement their Manufacturing needs, then they will pay for it. If the government is looking at AM to reduce Sustainment burden, then they will pay.”

• “Each should pay for their own; however, there should be good standards and guidelines to make this process easy and quick.”

• “If the military additive manufactured a part, let’s say at a remote base. We would have little choice but to do the qual/cert ourselves at the forward base where it’s printed.”

• “The government is paying for it one way or another, whether done distinctly or rolled up. Especially if it is an engineering change, which many AM-produced parts originally made via more traditional means would be, or if it were up front by design, the government would be doing it. After all, the government is determining whether the part is viable.”

• “Dependent on the type of qualification/certification. If this is airworthiness, then it is completely different story.”

• “If a new design AM part produced in-house by the government on government equipment, then of course the required qual/cert would be paid by the government.”

• “In-house is government in this case, so obviously we’d pay.”• “Only ones who can ... unless the producer is the OEM [original equipment

manufacturer].”

Additional comments from industry:• “Qualification/certification is part of the design process so the cost is paid for

by the customer.”• “It would have to be built into the selling price paid by the government with

any nonrecurring costs from industry recouped over the quantity of parts purchased.”

• “We typically do exactly this – paid for by the government and they own all data and rights at the end of our efforts.”

• “End customer generally pays for certifications. They tend to see the largest benefits in price reduction.”

• “The government has unique specs for the parts that they procure. Quals and certs would be based on their standards and specs.”

• “The customer pays for the qualification of their part.”• “As the government would own the final timeline, I would expect that the

qualification and final certification would be up to the end user.”

Question 3 discussion: Both the government and industry participants agree that the government would pay for the qualification and certification costs to produce a part-in house, 86% and 90% respectively. The cost of qualification and certification of a part is an expensive aspect of a business model; therefore, determining who would be incurring the cost was highly important. The business models are developed from the industry’s perspective, thus the qualification and certification costs are not included.



Question 4:Consider the situation where the government decided to produce spare parts organically (i.e., at a depot) using additive manufacturing instead of purchasing the spare part. If a supplier provided build files for government to additively produce parts in-house, what type of services would you expect to be included in the selling price? Check all that apply.

Additional comments from government:• “A lot of the above are standard engineering services.”• “Being able to manipulate the files is critical to fabrication.”• “Engineering support as needed. Doesn’t necessarily have to be on site.”• “Not necessarily a Help line, but do expect some help if necessary to understand

the build file and the process it was intended for.”• “Some of these items should be priced options, such as the Build process

update. I can only imagine that would be needed if there’s a configuration change to the machine, or if the machine itself goes obsolete.”

• “Dependent on Criticality of Item or Difficulty of item.”

Additional comments from industry:• “Re-design and update of file types may indeed be involved and offered as a

service at additional cost TBD.”• “[Additional services:] Prototype development and first article testing services.”• “Depends on the revenue stream. If a single purchase or nonrecurring, then

no services after initial delivery. If recurring fee, then services to maintain and upgrade product. Similar to software licensing agreements.”

• “NO guarantees the AM part will meet or exceed the durability of the current manufacturing process (no adverse residual stresses that could cause distortion or premature fatigue).”

• “NO guarantees the finishing processes will be identical and not adversely affect dimensional, corrosion, or fatigue and fracture characteristics.”

• “ESA [Engineering Support Activity] responsible for re-certifying the AM production part.”

Question 4 discussion: The question was formatted to allow participants to select all types of services to be included in technical data package cost. Eighty-nine percent of the government participants selected secure transmission as a service and 65% of industry participants. Secure transmission is a critical component to both industry and government members as the technical data package is being transacted digitally and the protection of these files are priority. Only 15% of industry participants believe build process updates should be included in the TDP price, assuming the machines will not be updated frequently. Twenty-five percent of both government and industry participants said a 24/7 helpline would be included in the file price; the result is lower than expected. The digital services cost included in the business models reflects the results of the survey.

Question 5:How much would you expect to pay for a single copy of a digital technical data package (TDP) file that would allow the government to print a single part?

Government sample comments:• “This will vary drastically based on part criticality, needed OQE [outgoing

quality evaluation], and number of expected parts that will be printed.”• “It really depends on whether the part is metal or polymer, the print time,

material used, and the complexity of the part.”• “Needs to be in relation to the total ROI for the part and parts design, but

significantly lower than the part cost.”• “Highly dependent on the part, but probably more than $5K for any structural

or critical parts.”• “I think that depends heavily on the part size, material, complexity, post-

processing, and level of criticality.”• “All depends on the part, as price could vary greatly from a plastic door handle

on a vehicle to titanium aircraft part and the amount of engineering that went into creating the TDP.”

• “Dependent on size, material, testing required – $125 - $1500.”• “Depends on the cost of the part. I would expect an o-ring to be MUCH less

than a fuel pump for example.”•

Industry sample comments:• “Depends on the value of the part. I would at least charge profit margin.”• “Depends on the certification/qualification cost and the quantity of parts

ordered.”• “Every part/systems is different so we cannot quote a price without looking at

a specific part or system.”• “Depends if the part is designed or off the shelf.”• “There is no set price. It all depends on what resources it took to develop the

part and the associated TDP.”• “Depends on the complexity of the TDP.”• “...would need some definition of the type, size, material, performance of part.

The charge could be the entire range specified up to 20X more.”• “It would depend on part and nonrecurring investment.”• “More detail would have to be specified around the scope included with the

TDP to determine pricing.”• “Depends on the quantity... DRM would be in play.”

Question 5 discussion: Only 10 government and two industry members selected a price range for the TDP. The other respondents selected “Other” and commented. Some of these comments are shown above. Both government and industry respondents recognized there are multiple considerations for the TDP price that will reflect the complexity, criticality, and other specifications of the actual part. Since the survey did not give an example part, the participants could not conclude or give a range of the TDP price. In the age of online data purchases for software, music, and movies, it is a positive sign for negotiations that both parties recognize that TDP data can be very valuable. Many factors contribute to the price; therefore, business models will reflect the uncertainty and the price will be varied.



The following two questions, Questions 6 and 7, were only addressed in the industry survey.

Question 6: How is profit determined in spare parts production?

Responses for “Other”:• “We do not make profit on parts – just on our services.”• “Unlike other industry (auto & appliance) after-market parts, the defense

industry is highly volatile. The defense supply chain needs to survive with other markets during the (extended) lean years. A predictable defense market for legacy parts is an important part to maintain the workforce skills and supply chain ecosystem.”

Question 6 discussion: The business models were originally developed using a return on investment to obtain profit in spare part production. After conducting the survey, 75% of the participants concluded a profit margin added to the cost to produce the part is used to achieve a profit in spare part production. Therefore, the profit margin business model was added to the analysis. ROI profit margin business model is still included in the analysis, but the profit margin added to the cost will be the focus of the business model analysis.

Question 7: How many employees does your company have?

Question 7 discussion: The majority of industry participants were either from small companies (<50 employees) or companies with greater than 500 employees—40%of each.

Author Biographies

Ms. Ashley Totinis the project engineer for America Makes, the National Additive Manufacturing Innovation Institute. As project engineer, she supports technical programs, education and workforce development efforts, and the innovation facility. Prior to joining America Makes, she was the engineering project manager for the Youngstown Business Incubator’s (YBI) Advanced Manufacturing Team where her efforts focu sed on cos t model i n g a nd bu si ness ca se a na ly si s pertaining to additive manufacturing. Ms. Totin received a BE and MSE in Industrial and Systems Engineering from Youngstown State University with a minor in Mathematics.

(E-mail address: [email protected])

Dr. Brett Conneri s d i rec t or of t he A dva nced Ma nu fa c t u r i n g R esea rch Center at Youngstown State University. Prior to joining the manufacturing engineering faculty at YSU, Dr. Conner was a U.S. Air Force officer, a defense contractor, and a research and development metallurgist at Alcoa. Dr. Conner received his bachelor’s degree from the University of Missouri in Physics, and masters and doctorate from MIT in Materials

Science and Engineering.

(E-mail address: [email protected])

Documents

for Maintenance and Sustainment - DAU 4 - Evaluating Business Models...Business model assumptions. Seven assumptions for the AM business model to follow: • The analysis is from a