Click for next page ( 2


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 1
INTRODUCTION Air Force sponsorship of manufacturing technology projects is often based on the hope that the results will not only benefit the original contractors but also will be transferred to other Air Force contractors. Some innovations seem to be readily adopted by other contractors, but others, though considered likely candidates for diffusion among contractors, are rejected for a variety of reasons. An understanding of those reasons and the process by which investment decisions are made will enable the Air Force to establish policies and procedures to enhance the likelihood of successful teabnology transfer to its contractors. At the request of the Air Force Systems Command (AFSC), the Committee on Computer-Aided Manufacturing studied three instances involving manufacturing research and development (~&D) projects completed under contract to the Air Force. The AFSC supported the projects with the understanding that detailed information about them would be made available without charge to other Air Force contractors. Each technology was considered by at least two firms other than the developer. In one case the technology was transferred; in the other two cases the technology has not been adopted by firms other than the developer, at least until this report was prepared in the summer of 1981. The committee has examined all three instances as case studies. From the evidence provided, the committee developed a model to describe the decision-making process used by potential adopters of innovations. Its objectives were to explain why attempted transfers of military- sponsored manufacturing technology succeed or fail and to propose changes in contracting procedures to increase the diffusion of such technology. In the following sections we describe the research procedures used, the framework for analysis, case study findings, and recommendations. Detailed case study reports form the three appendices. RESEARCH PROCEDURES A list of projects and the outcomes of attempted transfers was provided by the AFSC through the Air Force Materials Laboratory (AFML). Three cases were selected:

OCR for page 1
8 1. 2. 3. Hot isostatic pressing, Automatic assembly drilling, and Advanced composite tape-laying head. 1. General Electric investigated hot isostatic pressing (HIP) as a method to repair castings made of nickel, titanium, aluminum, and steel. Several vendors adopted the process. The conon~ttee research team contacted Howmet and TRW, who are adopters, and also studied intra-organizational transfer within General Electric. 2. Grumman had a contract to locate, precision drill, and countersink fastener holes by means of an automated drill-head mounted on a computer-controlled gantry. Grumman is currently using its automatic assembly drilling machine at a low rate of production. Though offered the machine, Fairchild and General Dynamics rejected it. Northrop discussed buying one but so far has not. The AFML considers the outcome of this case to be a failure. This judgment is based on a limited interpretation of transfer. It does not take into account instances where firms used knowledge of the Grumman work as a foundation for advancing the technology. 3. General Dynamics was one of several firms under contract to develop manufacturing methods for composite production integration equipment. The only advanced tape-laying head model currently in steady production at General Dynamics is a prototype machine. Ingersoll-Rand built a more advanced version for General Dynamics, but General Dynamics claims it has not worked reliably. Grumman rejected the General Dynamics approach in favor of one designed by LTV. McDonnell-Douglas might supplement its current broadgoods approach with a tape layer in the future, but at present no other company has adopted the General Dynamics concepts. The AFML considers the transfer of the tape laying head to be a failure. Each case includes an originating organization and at least two potential adopters. All told, eight firms have had an active part in the three cases. Consultants to the committee, Margaret Graham of the Harvard Business School and Clint Stanovsky of the Massachusetts Institute of Technology, interviewed key individuals--developers, evaluators, and decision makers--at each firm. The parties interviewed are shown in the table on the following page. FRAMEWORK AND ITS THEORETICAL UNDERPINNINGS Figure 1 is the decision-making framework developed for analyzing the empirical information. It shows decisions to be made before adopting an innovation, considerations that enter into each decision, and the research questions raised at each decision point. In general the committee conceives of the process as one or more dec~sion-makers weighing the perceived risk and perceived leverage of an innovation and comparing the outcome to other alternatives. 2

OCR for page 1
Case Study Interviews Firm Interviewees Status* General Electr ic Evendale, Ohio Peter Bailey Ernest Keraz in ick Ken Stalker Or iginator of 1 Howmet Bill Freeman Adopter of 1 Whitehall, Michigan Don Preston General Electric Lynn, Massachusetts Steel Irons Adopter of 1 TRW Jack Alexander {now at Precision Castparts Corp.) Potential adopter of 1 Grumman John Huhner Originator of 2 Bethpage, Long Island Carl Micillo Potential adopter of 3 General Dynamics James Ashton Potential adopter Forth Worth, Texas Grant Davis of 2 Wendal1 Eliot Orig inator of 3 McDonnell-Douglas Terry Howick Potential adopter St. Louis, Missouri Paul Meyer of 2 Potential adopter of 3 Northrop Don Stansbarger Potential adopter Los Angeles, California of 2 Potential adopte r of 3 *1. Hot isostatic pressing 2. Automatic assembly drilling 3. Advanced compos ite tape-laying head 3

OCR for page 1
c`. At a: o - o o - :~: c' A := v c c A: o := A ~ At A ~ ~ o o ~ - ~ v - - ~ - o ~ ;~: v, ~n ~ ~ ~: v ~ v ~s ~- =: ~ ~ ~-) u, <:: z o ~ :~ z ;- o ~ c~ o ~: - Z E~ o - : z; ~ n o c~ t~ - - - - - ~ o - ~ cL, a: o 4 o ~ 0 a ct O a, ': c, ~ v s~ 0 0 c, ~ ~ s" c~ ~ v v ~ 3 ~ a, c~ r~ e~ 0 ~ ~ E 3 ~ ~ C u' E c~ s~ au 1 ~V ~ ~ S" ~ ~ D O ~n ~ ~ 0 O ec E E a tJ 0 E c~ cO~ ~ ~ ~ ~ D c s~ ~ E . ~ 0 ~ ~ 0 c C 0 s" a, ~ ~ ~ E 3 P-= ~ ~ , o 1 e a, ~ 0 c: v ~o ~e c, 5 o 1 V t0 CO C r~ U: 3 o~ C`- _ ~ S~ _4 3 V ~ ~ ~ ~n eD C~ C . 3 u: 0 O C ~ - . 0 c a~ l ~ ~ -~: c a, ct u E ~ s~ co ~ ~ C c, ~ ~ 0 _ v "= ~ o, C: a~ s 0 ~ O <: o~ ^ ~ ~ o~ ~ ~ ct ~ ~ c`. c c ~ c ~ 0 ~ ct c 3 t~ u ~ c ~ 0 ~ a~ ~ c u, ~ ~ ~ c ~ tJ 3 C - 3 3 O ~ - 4 C ~ O E ~ a, ~ e~ ~ a, v 3 ~ O a u: ~ s" 0 3 ~ ~ ~ c~ 3 0 ~ c ~ ~ ~ ~ C15 D 1 ~ ~ ~ ~ ~ I 0 ~ ~ ~ ~ 0 c, eO 0 {~, ~ ~ z 0 ^ ~ ~ c c, ~ ~ 3 0 1 L' ~ ~t C~ z ~ 0 ~ oo ~ ~ ~ ~ c =: c~ _ CJ S" . E 0 C O s~ 0 ~ O z ~ ~ C a) ~ 0 a' ~ _ 0 ~ c: c~ s :, Q) _ s c~. e%. v 0 ~ 0 C ~ P. v a' 0 5 Fi 6 1 ~ ~ `: 0 c 3 0 0 0 0 u~ u~ u 0 ~ ~ ~ ~ v ~ ~ v X cc u, Z ~ ~ ~ c ,~: :d ~ ~c c,o ~ c`. v ~ ~ ~ ~ ~ ~ ~ ~ u, .,~ ~ 3 ~ s" z 0 ~ c) ~ v S v v cr X ~ X ~ 0 ~ 0 ~ ~ v v ~v u, 0 3 ~ 3 c ~ u, c 0 ~ 0 ~ e~. , J X U C E n C ~ 0 u: cu 7 ~U JJ cn ~a 3 _ ~ n o U~

OCR for page 1
The initial decision to consider adopting an innovation arises from recognition of a need, combined with recognition that a solution might be technically feasible. The decision to pursue the possibility or not will be based on perceptions of leverage (payoff relative to cost) and risk, tempered by perceptions of complexity and applicability of possible solutions. Research questions for this portion of the decision-~king process are: Was awareness of technical possibilities a significant help in the decision to pursue an innovation? A significant hindrance ? Was need a major determinant? Were alternatives considered? Was the specific innovation rejected in favor of an alternative? To what extent did the decision-makers proceed because of low risk? Because of high leverage? To what extent did the decision-makers proceed despite high risk? Despite low leverage? Was the decision making explicit? As the dec ision-maker approaches the selection of a particular option: . Were leverage and risk re-evaluated? Did these become more or less important as implementation became imminent? The final portion of the framework shows the decision to implement an innovation or not. The success or failure of an implemented innovation will provide feedback to future decisions about investments in new technology. Questions to ask at that time are: Did the attempted innovation succeed or fail. . . - because of (despite) accurate or inaccurate assessments? - because of (despite) determined or half-hearted pursuit? The committee does not address the last set of questions; its analysis ends with the decision to implement or not. In general this report seeks to determine the relative importance of each of the following: Awareness of technical possibilities, Risk reduction, Perceived need, and Leverage (payoff in proportion to cost). A range of interdisciplinary innovation literature has bearing on the characterization of the parties and the techniques in question. The industrial buyer behavior studies, for instance, apply diffusion research to marketing. Webster and Winds identify five 5

OCR for page 1
organizational roles {deciders, influencers, buyers, users, and gatekeepers) and four key determinants ~ individual factors, interpersonal factors , organizational factors, and environmental f actors) . With these, they correlate several decision stages in the buying process. Baker2 offers a less complicated model that weighs the importance of different factors in the buying decision. His approach teas the virtue of rank ing the importance of many of the factors identified by Webster and Wind. While his attempt to weight these factor s with preci s ion i s somewhat suspect, the notion of crude rank ing makes sense. A major factor that distinguishes the Air Force from other users of industrial innovation is its non-prof it status . Transfer of medical technology, then, may be analogous. Gordon and Fisher, in The Diffusion of Medical Technology, 3 sketch an excellent methodology that is similar to those cited earlier but emphasize the primacy of per formance (or ef f icacy in the case of drugs ~ for non-prof it enterprises . This study differs from most previous research on diffusion in that it seek s to explain not only the adoption of technology but also failure to adopt. This approach requires careful definition of the potential set of adopters. Another unusual aspect of this study is the attempt to characterize each innovation according to its chief technical attributes, as well as by the customary set of characteristics common to all industrial products. 4 White and Grahams characterize a technology according to the core concept, the embodiment, the operating characteristics, and the market characteristics. The concept of a new technology may be distinguished from its embodiment as a candidate for transfer. Not all successful transfers require that the embodiment be adopted; in some cases transfer of the concept is all that is intended or desired. ANALYSIS OF CASE STUDY FINDINGS Case studies are presented below in terms of the decision-making process at the originating f irms and at the potential adopters. The cases are described in greater detail in the appendices. HIP Casting Consolidation Technology Originators Both General Electric Evendale, which was the AFML contractor, and Howmet Turbine Components Corporation were originators of the hot isostatic pressing (HIP) casting process. Their decisions to pursue the technology were based on different planned applications. General 6

OCR for page 1
Electric, as an engine builder for military aircraft, specified in its designs castings that used HIP and also used the process to salvage unacceptable castings that bad been received from vendors. Bowmet and other parts suppliers used the HIP casting technique to produce products and to cut down on the number of castings rejected because of unacceptable poros ity . The HIP casting concept involves the application of high temperature and pressure to metal parts that have already been cast. The embodiment involves use of an autoclave to accomplish this application. The contract that General Electric performed for the AFRO provided for the use of HIP casting on metals used in airplane engine construction. The basic concept had already been demonstrated for aluminum castings by Alcoa when General Electric first experimented with HIP for casting other metals. Battelle's publicity for the techn igue prompted both General Electric and Howmet to consider further development programs. For both companies the original benefit of HIP appeared to be casting repair (scrap reduction for Howmet and salvage reduction for General Electr ic ~ rather than product enhancement. Dec i s ion to Pur sue General Electric's Evendale operation was responsible for investigating all relevant forms of manufacturing technology, and i t was Evendale that decided to pursue HIP casting. For General Electr ic, low r isk appear s to have been the important motivating f actor in the decision in two respects . First, General Electr ic was able to do its first experiments using an autoclave that was already available for its nuclear work . Second, and critically important, the AFML was willing to fund the project as soon as General Electric had brought the potential benefits and generic appeal of HIP casting to the attention of Air Force off icials. For Howmet low risk also played some part in the decision to pursue the technology, because Howmet was able to send early castings to Battelle to be processed in its autoclave. But perceived payoff seemed to be a stronger motivator for Howmet from the outset. The payoff was anticipated not only in terms of the tangible benefits that Howmet could realize if HIP casting reduced the scrap rate, but also in terms of the enhancement of Howmet's image as an innovator in its industry. Accordingly, Howmet investigated the HIP casting process at its own expense, avoiding the requirements that accompany government support for development. The first stage of investigation quickly revealed to both companies that HIP casting offered larger benefits than they had perceived at first. The process not only repaired bad castings but improved good ones as well. For Howmet this meant a potential competitive advantage in its market; for General Electric it meant the 7

OCR for page 1
possibility of improved performance in its engines. The new information made the adoption decision less complicated for Bowmet and more so for General Electric. While the findings increased the leverage of HIP casting for Howmet, it meant that General Electric Evendale had to convince the eng ine designers and General Electric Lynn, which produced its engines , to specify HIP-cast components in new engine designs. Dec i s ion to Adopt The decision to adopt HIP casting was a gradual one for both companies . With old equipment that could be modif led, Howmet was able to test its commitment to the process before investing heavily in new autoclaves specifically adapted for the new purpose. General Electric could also adopt gradually, by testing HIP-cast parts as replacements and in prototypes before actually using them in new development engines. As of 1980, Howmet can be said to have adopted HIP casting fully, General Electric provisionally. In view of the enhanced leverage revealed in the early stages of investigation, Howmet committed substantial sums to purchase an autoclave for HIP in 1975. By investing at this time, Howmet anticipated demand and planned to develop a market. General Electric Lynn is still evaluating the test data relating to the use of HIP castings as replacements, but there are strong indications that it may soon specify HIP castings for the next major group of development engines. Other Adopters Other parts suppliers have adopted HIP castings without committing themselves to major investment. The existence of companies that will use the HIP technique for them has made this possible. TOW, for example, will continue to send out its HIP casting work until the volume of demand seems to warrant buying autoclaves for internal use. Here, leverage clearly outweighs risk as a motivating factor. Findings from the HIP Casting Case 1. The AFML has termed the hot isostatic pressing contract with General Electric a case of successful transfer. Clearly AFML announcements and conferences, as well as conferences and reports generated by private research, bave been important in diffusing the conceded. General Electric's contract seems to bave had little to do with diffusing the embodiment of the technique. If General Electric does decide to specify HIP castings in its next major engine design, it will be instrumental in transferring the embodied technique as well. 8

OCR for page 1
2. In the case of HIP casting, the form of the technological embodiment was a factor that promoted diffusion. Although autoclaves , require significant capital investment, they are separable, stand-alone pieces of equipment that can be used with minor modif ications for a variety of tasks. The existence of autoclaves made it possible for HIP casting users to try different applications without interesting heavily ahead of time; and once a commitment to adopt was made it was possible to invest gradually. 3 . An important aspect of the HIP casting process, which may account for its successful diffusion, is its benef its when included in designs. This aspect of HIP casting secured the support of influential design engineers and increased the size of the potential market, while a llowing a premium to be charged for the product f abr icated as a result of the technique. 4. Because of the design implications, the relationship of the ANAL contractor, General Electr ic , to the potential adopters is important in the diffusion process. All suppliers can be expected to adopt the process in some way if General Electric specifies the use of HIP for cast components for its new engine designs. The timing of this diffusion will be closely linked with major new engine contracts. For all but conscious pioneers like Howmet, major new contracts are likely to trigger the adoption of the new technique. Automated Assembly Fixture Or illing The motivation for automated assembly f ixture drilling was a desire to automate labor-intensive and monotonous tasks in airframe assembly. This need had been defined and promoted in several Air Force conferences in the early 1970s. Automation in assembly would not only reduce cost and improve consistency, but it would reduce dependence on trained manual personnel. With the recent appearance of minicomputers capable of being operated on the shop floor, the enabling technology was at hand. Both Grumman and General Dynamics chose to pursue concepts related to automated drilling. The Air Force Materials Laboratory chose Grumman as its contractor to develop and demonstrate automated drilling in preference to General Dynamics, which was also competing for the contract. The Grumman approach was preferred because it added a scanning mechanism as a locating device in addition to the computer control cuff the dr illhead . Or is inator The factor that f irst motivated Grumman to pursue the automated drilling technology was its perceived leverage. While Grumman lacked a high volume airframe contract to which the technique could be applied in the near term, Grununan's Advanced Materials and Development 9

OCR for page 1
Group had a standing mandate to explore all potential manufacturing cost-reduction opportunities. Grumman saw in the technique a variety of potential short-term payoffs dumb as chances for quid-pro~uo subcontracting, royalties, and demonstration funding from the Air Force for technology that had generic appeal to the industry at large. Demonstration of the automated drilling technique entailed relatively little risk for Grumman. It was able to show feasibility for the equipment by drilling production panels in Plant Twelve where the equipment bad originally been developed e Later it was able to move the same prototype equipment to the A-6 assembly line. Volumes on the A-6 program were low enough that the prototype automated assembly fixture was adequate for the task without modification. The same piece of equipment could be used to perform the Air Force contracts, enabling Grumman to evaluate the economics for both the A-6 and the A-10 parts. Adopters Since all airframe manufacture requires drilling numerous holes, the Grumman device was viewed by Grumman and the AFML as highly generic and potentially transferable to all airframe manufacturers. Immediate potential adopters were those companies tooling up for new contracts in 1974, Fairchild and General Dynamics. The AFML suggested that Grumman demonstrate the technique on an F-16 par t produced by General Dynamics. In the view of the AFML, then, General Dynamics was the designated adopter. Other airframe manufacturers monitored the development of new techniques but were unlikely to adopt new tooling in the middle of a program. Despite the assessment of its general quality, the Grumman device has not transferred to any other company as of 1980, though several companies have given it favorable evaluations that may result in transfer in the future. General Dynamics was one of the first companies to be given a demonstration of the automated assembly fixture drilling, even before Grumman received AFML funding to demonstrate it. General Dynamics had also been pursuing automated drilling, using its own research money. Its loss of the development contract to Grumman naturally affected its evaluation of Grumman's approach, but the ultimate decision not to adopt the Grumman device was motivated first by high perceived risk and secondarily by questionable leverage. The perceived risk in the Grumman device der ived f ram two sources- - oubtful appl inability and poor relations between the two companies. Leverage appeared low, not only because Grumman proposed to charge substantial royalties, but a lso because the egu ipment would be costly to repl icate at General Dynamics in view of the poor communications between the two companies. Grumman had tried unsuccessfully to interest Cincinnati Milacron in building the machine, and General Dynamics was unwilling to rely on Grumman drawings as the basis for the transfer. In the end General Dynamics chose to reject both the Grumman device and its own 10

OCR for page 1
earlier approach to automated drilling in favor of a robotic wing driller . This device reflected General Dynamics's preference for lower cost, less deaf icated equipment . McDonnell was not tooling up when Grumman first demonstrated its automated assembly fixture drilling system, though its engineers thought the concept had high potential leverage. After watching Grumman ' s demonstration, McDonnell seriously considered investing in the technology. Evaluation of the Grumman approach ultimately revealed poor applicability without extensive modif ication . There was high perceived risk in replicating equipment that was not built by a machine tool maker. In the end McDonnell also chose to develop its own equipment, but, unlike General Dynamics, it chose to adopt certain Grunt an concepts--in particular, scanning for accuracy of hole locat ion . Fairchild came closest to actual adoption of the Grumman equipment, signing a lease agreement in 1976. In the end insufficient leverage prevented the transfer there too, but in Fairchild's case the problem was one of timing. Had the Grumman demonstration occurred when Fairchild was tooling up for the A-10 instead of a year later, Fairchild would probably have adopted the technology. In 1977-78 Northrop considered buying a wing drilling system, with the understanding that it would cost S250, 000. Northrop' s manufactur ing processes group conducted a feasibility study that showed a marginally acceptable payback. But when Grumman raised the price to S1.2 million, Northrop could no longer anticipate suff icient leverage in adopting the system and rejected it on economic grounds. Findings from the Automated Fixture Drilling Case 1. The AFML has labeled the Grumman device a case of failed transfer, yet it may still attract adopters. The Grumman concepts for automating drilling, especially the use of a scanning device for hole location, have already been transferred, even though the complete embodiment has not. The demonstration of the Grumman device clearly stimulated at least one company, McDonnell, to look at automated drilling in assembly for the first time. 2. Transferring technology into an interrelated system such as an assembly operation is bound to require some adaptation. me amount of adaptation depends on two factors: the similarity of the products being assembled and the similarity of organizational design and manuf actur ing ph i losoph ies . Both McDonnell and General Dynamics found that the Grumman device needed significant adaptation for their needs, in part because of their h igher volumes of production and in part because of the different types of drilling required. Fairchild and Northrop have production volumes more comparable to Grumman's, and their manufacturing philosophies are similar.

OCR for page 1
3. Relations between the companies involved in a transfer have a profound influence on the success or failure of adoption. If past dealings beve been good or if the companies are currently involved in joint work, such as a subcontracting arrangement, then the process of transfer is aided. By contrast, if there is a history of previous conflict among the parties to a potential transfer, the perceived risk of transferring and adapting a technology becomes high. Even very thorough reports and demonstrations contain only a fraction of the information and know-how required to transfer a complex embodiment. 4. The Grumman device would have been much more likely to spread in its embodied form if a machine tool builder had replicated the device. In cases such as General Dynamics and McDonnell, where adoption involves significant adaptation of the original embodiment, the costs of transfer may well exceed the cost of building equipment from scratch. If a machine tool company, with its know-how and warranties, were to produce the equipment, adopters might be willing to forgo some adaptations. The Advanced Composite Tape-Laying Head The advanced composite tape-laying head is designed to automate the process of fabricating laminated parts from advanced fiber composite tape. Composite materials (boron or graphite fibers in a resin base) are available in either tape or broadgoods form. Uncertainty concerning the format and cost of the materials has kept production technology f lu id over the past 20 years . There are numerous materials suppliers, and although there has been some standardization of widths, there are still many different combinations of material and adhesive systems, each with slightly different handling properties. With the additional problems of storage and poor shelf-life, the difficulties of settling on a stable production system become enormous. Whether in tape form or in broadgoods the material has to be dispensed, laid up carefully ply on ply, and cut accurately in any of several ways, not necessarily in that order. The concept of the tape-laying head is the automation of this process for the tape format, heretofore an intensely manual process. The tape is dispensed, its deposition controlled so there are no gaps or overlaps, and then it is sheared evenly and accurately. Originator When General Dynamics chose to pursue the automation of tape laying in the mid-1960s it seemed to all that the pr ice of composite materials would decrease and more composites would be used as a consequence. The company' s first prototype machine gained support f ram the AEON to develop an improved version built by Conrac. The series of AFML contracts that followed reduced the risk for General Dynamics; to interest in tape-laying automation but the real motivation

OCR for page 1
was perceived leverage. In the late 1960s General Dynamics planned to design a high-volume, low-cost f ighter aircraft , using some composite parts, and manual production costs were considered to be prohibitive. The firm projected a need for 15 tape-laying machines in the mid-1970s. Since all military aircraft were expected to incorporate composite materials in a few years, the AFML saw the automated tape-layer as highly generic. General Dynamics experimented further with the Conrac and other improved prototype heads. It adopted and modified its concept in a machine built by Ingersoll-Rand for production of the F-16 in 1976. In 1977 it began a further AFML contract to improve the tracking capability of the head, as well as to introduce flexibility as to length of strips laid down and versatility in cutting. Its latest AFML contract was designed to perfect the concepts for use in an integrated, fully automated composite production system. General Dynamics chose to stay with the tape approach, even after broadgoods became available, not only because it had already invested extensively in tape technology but because it saw tape as the lower cost approach (less waste, more versatility, fewer materials control problems). The company placed such importance on low-cost manufacturing that it was willing to limit the freedom of its designers if that were necessary. Since the time material suppliers made broadgoods available, a number of equipment makers that had previously focused on the garment industry have entered the aerospace market. Because of the large number of competitors in the field, the equipment builders have tended to custom design equipment--and charge custom prices. General Dynamics has not yet found an equipment maker to produce its most recent version of the tape-laying head at an acceptable price. Adopters The tape concepts that had seemed generic in the late 1960s when the AFML funded the early General Dynamics contracts were called into question when broadgoods became available. The broadgoods philosophy won enough converts to narrow the field of potential adopters considerably. Three different groups emerged: companies such as General Dynamics that stayed with tape, companies such as Grumman that adopted a hybrid philosophy, and companies such as Northrop that moved entirely into broadgoods. Grumman closely followed General Dynamics into the area of composites automation. It began investigating the automated tape- laying concept when General Dynamics demonstrated the Conrac. In 1969 Grumman projected a need for perhaps one-third of the volume projected by General Dynamics . Grumman evaluated the Conrac for its own use but rejected the General Dynamics embodiment in favor of its own mechanized 13

OCR for page 1
tape dispenser. The objections to the General Dynamics approach might be regarded as technicalities, but they reflected enduring differences in priorities. General Dynamics emphasized cost and volume; its primary concern was economic. Grumman insisted on various performance characteristics such as individual ply inspection and accurate cutting before it turned to cost considerations. As a result, Grumman chose to pursue other available options, including its own. It learned from General Dynamics's concepts but did not adopt them. Grumman moved closer to adopting a particular tape layer in 1974-75 when it foresaw a need to automate its composite production in order to manage the huge B-1 bomber stabilizer. Its need was again defined not so much in terms of cost as in terms of performance. Having evaluated the three leading tape layers, it chose LTV' s because it would lay up smaller individual pieces of tape than the one for General Dynamics. Grumman then pushed for defining an entire integrated composite production system and reducing cost on a system-wide basis . In 1975-76 the AFML funded Grumman ' s integrated laminating center, thus leading the firm permanently away from the advanced tape head concepts. When broadgoods appeared in the early 1970s, other companies reassessed their entire composite production systems. McDonnell had been track ing and evaluating the automated tape-laying concepts at each stage and had built its own more rudimentary equipment. McDonnell 's leadership in composites was based on sophisticated design, not manufacturing technology. Broadgoods seemed to offer more flexibility to designers. McDonnell opted for laser cutting as its ma jor production investment. In the end, therefore, compatibility with manufacturing philosophy became the key factor in McDonnell's non-adoption of the tape-laying device, and awareness was the trigger for adoption of the alternative system. This decision may be changed eventually. McDonnell could still adopt tape-laying equipment when it does such a high volume of composite parts and structures that a subsidiary tape capability becomes desirable to enhance flexibility. Northrop waited to pursue automation in composites until broadgoods were available. Its volume of composites was so low that risk reduction and awareness of technologies consistent with its philosophy were the decisive factors. Manufacturing flexibility is the main consideration for Northrop. Since broadgoods satisfy that requirement more than tape, the company does not consider itself a member of the class of potential adopters for a tape-laying head in the foreseeable future. Findings from the Advanced Composite Tape-Laying Case 1. The AFML has judged the advanced composite tape layer to be a failure not only because no other company has adopted the concepts but because General Dynamics has yet to put its most recent improved 14

OCR for page 1
version of the head into production. The problem is the difficulty of getting an equipment builder to produce the head at an acceptable price. 2. In a technology as fluid as composites technology, embodiments are extremely unlikely to transfer. Each attempt to - embody concepts raises new problems and the whole system is so unstable that embodiment in other than prototype form is prohibitively risky. Even concepts in this environment are more likely to stimulate further development than to transfer intact. 3. An added barrier to transfer is the reluctance of machine tool builders to become involved in an unstable technology without charging custom prices. 4. Individual manufacturing philosophies, consistent over time, play an important role in aiding or obstructing transfer. Even though General Dynamics and Grumman have very similar composite concepts, the differences in their manufacturing philosophies inhibit transfer. OBSERVATIONS AND CATIONS Aspects of Technology Whenever technology transfer is discussed, too little attention is generally directed towards the characteristics of the technologies. A f ew distinctions need to be borne in mind. First, it is important to distinguish between concepts and thei r embodiments. It is possible for concepts to transfer while particular embodiments--physical configurations of those concepts--do not. The reverse is also true. For each individual technology, therefore, it is important to decide whether the real value is in the technology concept or in its embodiment. A transfer should be j udged successful if the valuable part has transferred. Another aspect of the concept versus embodiment question is related to the category labeled generic. A striking feature of the cases treated here is that although the concepts judged to be generic frequently proved to be so, the particular embodiments often impeded their transfer. Some embodiments, such as that of hot isostatic pressing, are more permissive In this sense than others. While there has been concern about the amount of capital investment involved, perceived risk is related as much to flexibility, reusability, and adaptability as it is to actual risk. Yet a third aspect of a technology that must be noted is the uncertainty associated with a high rate of change. If a whole process area is changing as rapidly as composite production is changing, for instance, then to look for transfer of whole concepts, let alone of 15

OCR for page 1
whole embodiments, is to look for premature standardization. In a rapidly changing field, stimulation of new concepts may be the greatest contribution that an Air Force-sponsored project can make Aspects of Transferring Organizations The Committee on Computer-Aided Manufacturing, in it'; 1979 annual report, addressed technology transfer and the characteristics of participating organizations. That report distinguished between transfer to large sophisticated firms and to smaller and less sophisticated organizations, between transfer within the aerospace enterprise and to or from non-aerospace firms. The committee recommended in 1979 that the ICAM program take advantage of the potential role of hardware and software vendors and machine tool builders, that it stress communication between transferring and adopting organizations, and recognize that standards for systems design will be adopted more readily than computer code. 6 The case studies just completed tend to confirm those observations. Compatibility between the existing systems and Phil osophies of the parties to a potential transfer are a necessary but insufficient condition for adoption. Minor differences can be modif led, but the costs of modif ication and communication soon exceed the cost of in-house development for most system embodiments. This factor is one of the main reasons that machine tool builders often play an important role; the value they add as a neutral party reduces the urge to redo the embodiment. Rapport between originators and adopters significantly reduces the perceived risk in adoption as well as the costs associated with transfer . Companies such as Grumman, Northrop, and General Dynamics, which possess it&D-dedicated groups that routinely track process technologies are likely to pursue concepts as opportunities, without a high perceived leverage. However, clearly defined need and high leverage are critical to outright adoption by all companies. For Air Force contractors, outright adoption hinges almost always on a major new program, because that is when capital investments are made. Awareness of demonstrated capability in a particular new setting may help to reduce the perceived risk, but it is not enough alone to stimulate any activity other than consideration. Availability on the open market, however, may well lead to adoption because the machine tool builder offers ways of reducing the risk and buffering the uncertainty. The phrase awe are not machine tool builders,. beard so frequently among contractors, indicates what an important, if indirect, role machine tool suppliers have played in initiating or inhibiting the transfers that the Air Force has wished to encourage in the past. 16

OCR for page 1
Recommendations for Air Force Action The observations stated above have implications for Air Force practices that can or do influence technology transfer. Recommendations to enhance technology transfer among Air Force contractors follow. First, it would be useful for the Air Force, when considering awarding contracts designed to encourage transfer, to address the distinction between concept and embodiment in light of objectives. In many instances ache real objectives, carefully defined, can best be met by proof and demonstration of a concept. The Air Force should broaden its interpretation of successful transfer to recognize the benefits of transferring the concept. If a generic embodiment transfer is really required, the embodiment might be developed cooperatively or in cooperation with a machine tool builder at the start. If a technology has significant and apparent benefits when included in designs, there may be no need to fund development or demonstration as it is likely to be adopted rapidly in the natural course of events. Second, the Air Force should take into account the often decisive role of machine tool builders in transferring technologies. To gain their cooperation the Air Force might offer incentives that are not administered through the contractor. Third, guidelines for dealing with potential users ought to take into account their problems in adaptation. Better contracting procedures might help to establish the responsibilities of originators and adopters as a formal condition for funding. Efforts should be made to identify receptive users and to consider their needs early in a development program. Gaining adopters at an early stage, particularly receptive adopters with a stake in the technology to be transferred, could greatly increase the acceptance and use of the new technology. Recommendations for Additional Study The three cases are examples of hardware manufacturing technology. However, the framework for this study also applies when the embodiment of computer-aided manufacturing technology is software. As suggested from the case studies, it Is important to distinguish between concept and embodiment; in the case of software the embodiment would be represented by the code--either source code or machine code. The case studies suggest that computer-aided manufacturing technologies may transfer easily in concept, whereas the embodiment (code) may prove not to be generic or transferable at all. In these three cases it appears that a machine tool builder's active involvement will usually improve transferability. Analogously, 17

OCR for page 1
a software house or even a computer hardware house might be important to the transferability of computer-aided manufacturing technologies. This discussion suggests significant applicability of the present studies to issues concerning the transfer of computer-aided manufacturing technology. To test these inferences we recommend the study of one or more additional cases involving the transfer or potential transfer of a computer-aided manufacturing technology where embodiment is a computer code. As manufacturing systems become more complex and more integrated, transfers of hardware/software combinations will be increasingly common. We further recommend one or more case studies of the transfer of systems that, of necessity, involve such combinations. NOTES Frederick E. Webster and Yoram Wind, Organizational Buying Behavior (Englewood Clif fs , N. J.: Prentice-Hall, 1972) . 2Michael J. Baker, "Industrial Buying Behavior and Adoption, n in Michael J. Baker, ea., Industrial Innovation : Technology, Policy, Di f fus ion (McMillan: London, 19 79 ~ . . 3Gerald Gordon and Lawrence Fisher, eds. The Diffusion of Medical Technology (Cambridge, Mass.: Ballinger Publishing Co., 1975~. 4George Hayward, in Baker 1979, compares the attributes of a technology as perceived by different parties to a transfer transaction. 5George R. White and Margaret B.W. Graham, Chow to Spot a Technological Winner, n Harvard Business Review, v. 56, no. 2, Mar-Apr, 1978. 6Co=,littee on Computer-Aided Manufacturing, The Committee on Computer-Aided Manufacturing in 1979, Annual Report (National Academy of Sciences: Washington, D.C., 1980) pp. 20-21. 18