Tech Data Transfer

THE IMPACT OF TECHNICAL DATA TRANSFER PROBLEMS DURING A TRANSITION OF WEAPONS SYSTEM PRODUCTION BETWEEN NATIONS Michael E. Schaller

In the early 1980s the U.S. Army sought to replace its aging fleet of M102 105mm-towed with the British M119 under a Non-Developmental Item (NDI) acquisition strategy. This paper addresses the problems experi- enced, their possible causes and effects, and provides a list of lessons learned. Finally, recommendations are made to help future program manag- ers mitigate or avoid the problems experienced by the M119 program.

lthough the M119 program has Additionally, the Program Office was been hailed as a very successful restricted in its ability to mitigate some NDI acquisition, there were of the potential risks associated with problems encountered by the program TDP transfer. All in all, the TDP trans- office along the way. The most signifi- fer problem cost the program an in- cant of these was the transition of pro- credible amount of time and money. duction from the original manufac- All of the data presented here (with turer, Royal Ordnance, to the U.S. ar- the exception of Reese and Fowler, senals at Rock Island, Illinois and which are periodical articles) were Watervliet, New York. The major cause drawn from American sources. Neither of this transition problem was the trans- Royal Ordnance nor the government of fer of the Technical Data Package the provided input to (TDP). this work. Representatives from Royal The TDP provided by Royal Ord- Ordnance were contacted via facsimile nance was not, and never would be, but did not respond. One source did found acceptable under U.S. standards. agree to discuss issues contained in this

39 Acquisition Review Quarterly – Winter 1996

work based on a grant of anonymity. the LID requirements for light infan- The author was also provided multiple try (HQ, U.S. Army AMCCOM, 1985, examples, from multiple sources, of pg. 1). Over the course of the next five General Officer “meddling” and pres- months AMC evaluated 20 weapons sure in the M119 acquisition process. and eliminated all but four. It was from these four that the British Light Gun, the L119, was determined to be the FROM WHENCE IT CAME “best candidate for the LID” (U.S. Army AMCCOM, 1985, pg. 2). The In late 1983, the U.S. Army initiated ARDEC briefed these results to the a program to redesign the structure, CSA in May 1984, recommending the roles, and missions of its light infantry L119. divisions (LIDs). Within this redesign The Chief’s decision was to lease a of the LID, the decision was made to sufficient number of L119 howitzers for procure “a longer range, more lethal testing, and to develop new 105mm weapon” (Army Magazine, rounds for increased range and lethal- 1986, pg. 365). Importantly, the Army ity (HQ, U.S. Army AMCCOM, 1985, Chief of Staff (CSA) “also established pg. 2). After this initial testing was suc- an extreme sense of urgency for field- cessfully completed, the weapon was ing the light division” (U.S. Army type-classified in December 1985. Pro- ARDEC, 1987, pg. 1). The program di- duction contracts were prepared and a rection that devolved from this CSA licensing agreement between the U.S. guidance was to search for a and Royal Ordnance was negotiated that the Army could “field immedi- the following year (Armament and ately” (U.S. Army ARDEC, 1987, pg. Chemical Acquisition and Logistics 1). In terms of the M119 program, “im- Agency, 1994, pg. 1). mediately” was translated into a “must The licensing agreement was made have” fielding date in selection criteria necessary by the Army’s decision to that would fall within fiscal year 1987 purchase only a portion of the weap- (U.S. Army ARDEC, 1987, pg. 3). ons desired from Royal Ordnance, with In January 1984, Army headquarters the remainder being produced domes- tasked the Army Materiel Command tically. It authorized American produc- (AMC) to search the inventory of U.S. tion of the L119 and established roy- and NATO 105mm howitzers and de- alty payment procedures. velop a list of those capable of meeting The decision to produce the M119

Major Michael E. Schaller, USA is assigned as a contract specialist to the Army Re- search, Development and Engineering Center (ARDEC) at Picatinny Arsenal, New Jersey. He received his B.S. degree from the United States Military Academy at West Point, New York and his M.S. degree from the Naval Postgraduate School in Monterey, California.

40 Tech Data Transfer

domestically was based on two factors. IMPACT OF THE TECHNICAL The first, maintainability, stressed con- DATA PACKAGE cerns about the availability of spare parts and the lack of control over an A technical data package defines the offshore source. Second and conceiv- system’s design configuration and the ably more important, there were na- production, engineering, and logistics tionalistic considerations; specifically, support procedures required to ensure the maintenance of the mobilization the system’s adequate performance. base. Each had an impact on the deci- The TDP consists of all applicable tech- sion (U.S. Army AMCCOM, 1987). nical data, including drawings, quality Based on these and other factors, the assurance provisions, and packaging Government opted to perform produc- details (DSMC, 1991). tion at Watervliet Arsenal, which would As part of its licensing agreement manufacture the cannon assembly, and with the U.S., Royal Ordnance agreed Rock Island Arsenal, which would pro- to provide the Americans with a TDP duce the trail assemblies. The two that was “sufficient to manufacture in pieces would be assembled, and the the U.S.” (Armament and Chemical howitzer completed, at Rock Island. Acquisition and Logistics Agency, 1994, These two arsenals, then, had to pre- pg. 1) and which “consists of all the re- pare their facilities for the gradual tran- corded ‘know-how’ required to manu- sition of production from Royal Ord- facture, assemble and test...the L118/ nance. L119 gun” (Nathan, 1995, January 30). The license agreement cost 1,150,000 What Royal Ordnance actually pro- pounds sterling for 145 complete how- vided was much less: an archival set of itzers, 20 carriage assemblies, 15 trail drawings, a set of manufacturing draw- assemblies, additional parts and ings (which showed in-progress draw- equipment from the U.K., and the ings, some gage and inspection draw- royalty fees and TDP required for ings), an il- lustrated subsequent U.S. production (U.S. ...the TDP as delivered parts cata- Army AMCCOM, 1987, pg. 1). The was “’archival’, grossly production contracts were signed in log (similar inaccurate, and missing July 1987, with Royal Ordnance oper- to our -34P essential manufacturing ating off what was considered to be a technical data” warm production base. Royal Ord- manuals), nance delivered the first production and a Final Inspection Record (Nathan, guns to the U.S. for Production Verifi- 1995, January 30). While all of these cation Tests in early 1988. Production documents are valuable in and of them- began in America in fiscal year 1988, selves, they did not satisfy the U.S. with the first howitzers coming off the Government’s requirements for TDP line in October 1990 (Reece, 1991, pg. content and accuracy of drawings. 718). According to the original schedule, the technical data package was to be de- livered in January of 1986. It was not

41 Acquisition Review Quarterly – Winter 1996

actually received, however, until August The cost of this domestic fix was $3.0 of that year, due to problems that Royal million (Armament and Chemical Ac- Ordnance experienced in collecting and quisition and Logistics Agency, 1994, assembling the required information pg. 1). The end result was a TDP nearly (Nathan, 1994, pg. 18). Additionally, $7.75 million over budget and more the TDP as delivered was “‘archival’, than three years’ late. grossly inaccurate, and missing essen- These problems highlight the poten- tial manufacturing data” (Armament tial difficulty in dealing with sources and Chemical Acquisition and Logis- other than those routinely involved in tics Agency, 1994, pg. 1). production for the U.S. Government. The engineers at Rock Island and This is not to say that similar problems Watervliet recognized that this TDP never occur with domestic manufactur- was of little use to them and returned ers. Rather, the problems were signifi- it to Royal Ordnance for rework. Royal cantly exacerbated by Royal Ordnance claimed in response that the Ordnance’s inexperience with the Technical Data Package met the re- American “way of doing business.” quirements, One of the critical requirements of “Royal Ordnance had but agreed to an American technical data package is no idea what an Ameri- fix it for an ad- that it provides the information re- canized Technical Data ditional cost of quired by manufacturers to “produce Package looked like” $4.75 million. to [the] TDP with stringent configura- The Americans tion management requirements” in the program office felt “in a bind” at (Nathan, 1994, pg. 1). This requirement this point: Royal Ordnance had failed caused a significant portion of Royal to comply with the licensing agreement Ordnance’s TDP problems. The com- and should fix the Technical Data Pack- pany produced the L119 in its own plant age at no cost, yet there was severe pres- using a “fit at production” philosophy, sure to get the howitzer into the sys- so that the accuracy of drawings used tem quickly. As a result, the program on the production floor was less criti- office opted to avoid lengthy litigation cal. However, Royal Ordnance had dif- by paying Royal Ordnance to rework ficulty putting this process on paper. As the TDP. one member of the project team stated, This revised TDP still failed to meet “Royal Ordnance had no idea what an requirements in August 1987 (Nathan, Americanized Technical Data Package 1995, January 30). Again, the problems looked like” (Nathan, 1994, pg. 1). centered on the actual quantity and Furthermore, the “British Technical detail of information being provided. Data Package also had a substantial Schedule slippage, coupled with the amount of sole source or proprietary delay engendered by Royal Ordnance components, which is unacceptable in in modifying the original TDP, a U.S. Technical Data Package” prompted the project office to have this (Nathan, 1994, pg. 2). The British pro- version fixed at ARDEC and at the ar- curement process does not require senals in Rock Island and Watervliet. competition. As a result, their system

42 Tech Data Transfer

has no need for the TDP information what the U.S. government would con- which is typically used by Americans to sider a TDP. To achieve that level of facilitate competition among different accuracy and detail, Royal Ordnance commercial sources or, alternatively, to had first to update the TDP they were produce the component in a U.S. Gov- using in line with the howitzers they ernment arsenal. were producing. This required that they The real issue had little to do with revise virtually every drawing to reflect the TDP itself, but rather with the dif- the waivers, deviations, and engineer- ference in the production philosophies ing changes already approved on the of the Americans and Royal Ordnance. shop floor, then implement a configu- Essentially, at Royal Ordnance each ration management and status account- howitzer was built individually, with ing system to ensure that any subse- pieces machined to fit each weapon quent revisions were recorded on the regardless of design drawings. These spot (Nathan, 1994, pg. 2). This pro- production floor changes were seldom, cess, undertaken by Royal Ordnance if ever, reflected in the technical draw- with extensive U.S. help, took well over ings included by Royal Ordnance in the a year. Once completed, the technical TDP it provided to the U.S. Essentially, data products provided by the the TDP failed to reveal the actual pro- company’s manufacturing element im- cess followed in manufacturing the proved significantly, although they re- L119. mained below U.S. standards. Given no requirement for changes to The impact of these TDP problems be tracked or reflected on drawings, was enormous, driving program cost configuration control of the L119 was $24 million above budget and delaying also a problem. In fact, with no stan- initial fielding by more than three years. dard manufacturing process and no The TDP itself cost nearly eight times approved design, it seemed nearly im- the amount originally planned. possible that two identical howitzers In the absence of a good TDP, the could roll off the production line. initial 1984 estimate for retooling Rock Island and Watervliet arsenals to pro- duce the M119 was $8 million, based FIXING THE PROBLEMS on historical data from production of the M102 howitzer. This history failed Based on the differing views of pro- to provide an accurate projection of the duction and configuration management requirements for the M119. In early between the U.S. and Royal Ordnance, 1985, the first revised estimate in- it isn’t surprising that the TDP provided creased tooling costs to $10 million. An- by Royal Ordnance would fall short of other refinement, which took place just what U.S. manufacturers required to prior to the receipt of the first Royal produce the M119. Ordnance TDP, raised the estimate to The drawings provided by Royal $13 million (Nathan, 1995 [January 30], Ordnance in August 1986 were really pg. 1), or $4.75 million each for nothing more than a rough draft for Watervliet and Rock Island in fiscal

43 Acquisition Review Quarterly – Winter 1996

year 1987, with roughly an additional drawings to make them suitable $3.5 million for Rock Island alone in for U.S. arsenal approval (Fahey, fiscal year 1988 (HQ, U.S. Army 1994, pg. 2). AMCCOM, 1987, pg. 3). However, the eventual receipt of the TDP, and the To expedite the overall effort, the subsequent revisions made to it, re- concurrent engineering team at Rock sulted in a final estimate for tooling Island forwarded the ‘Americanized’ costs of $23.3 million. versions of individual Royal Ordnance Army and AMC staffs approved TDP drawings to the manufacturing these funds, and production tooling floor as they were finished. This pro- began in March 1990 (Nathan, 1995, cess, which was both time and man- January 30). As a result of TDP prob- power intensive, produced a TDP that lems, tooling costs for the arsenals were was “not an optimal Technical Data almost three times the original esti- Package, but was a Technical Data mates. Package that the arsenals could pro- However, because production tool- duce to” (Fahey, 1994, pg. 2). ing (and production itself) could not As of the summer of 1995, there is begin in the U.S. until a usable TDP no competitive TDP available. Produc- was developed, the timeliness of Ameri- tion is still being conducted from Ord- can production was threatened. The nance Drawings produced at Rock Is- transformation of the production lines land Arsenal. at both arsenals depended on the abil- The problems with the M119 tech- ity of their engineers to estimate and nical data package during transition of forecast equipment and material re- production to the U.S. provide an im- quirements. This estimation process, portant source of information for fu- usually based on some form of techni- ture program managers. cal drawing, is critical to a rapid transi- tion. Facing an ever increasing amount 1. Technical data transference is criti- of pressure, “(T)he arsenals could not cal if production transition is to be afford to wait for an Americanized effective. Virtually every problem Technical Data Package in order to start associated with the domestic pro- production” (Fahey, 1994, pg. 2). In- duction of the M119 stems from stead, the process went ahead using the inaccuracies and problems data gathered through a concurrent en- with the Technical Data Package. gineering effort at Rock Island. In this No significant problems were ex- process, perienced with the actual physical reconfiguration of the arsenals to (A) concurrent engineering team do the production. Once the re- (Arsenal production, ARDEC en- quired information was available, gineers, production planners, the arsenals functioned as they quality control and product assur- were supposed to. In this case, ance and logistical people) [got] Royal Ordnance was not necessar- together to review and mark up ily unwilling to provide accurate

44 Tech Data Transfer

technical data. In fact, due to the office into an untenable position structure and process by which in terms of contract clause en- they had been producing the forcement. In this case, the buyer L119 for the U.K. Army, they ‘needed to buy’ more than the were unable to provide an seller ‘needed to sell.’ As a result, American quality TDP. the program office had difficulty forcing Royal Ordnance to live up 2. Foreign suppliers may or may not to the agreements of the contract; understand our acquisition prac- it was faster to concede and pay tices. Clearly, Royal Ordnance did the extra money than it was to fight not. Issues which are peculiar to it out. U.S. acquisition process in gen- eral, and to TDPs in particular, 4. Trying to fix something after the such as proprietary or sole source fact is hard to do. Once the con- information restrictions, the level tract was awarded to Royal Ord- and degrees of accuracy for tech- nance, it became extremely diffi- nical drawings, and rigidness of cult to ‘force’ them to change and configuration control can intro- do things our way. This was the duce serious problems into the case with the discrepancy over the acquisition cycle if not handled original Royal Ordnance-deliv- properly. ered TDP with regard to its com- pliance with the license agree- 3. Buying in a ‘rush’ is dangerous. ment. Royal Ordnance claimed With the selection of an existing compliance, and it would have system, the Army hoped to pro- been extremely costly in terms of cure a weapon system in less time time and money to force them to than that required for a full devel- do something which might possi- opment. The use of a Non-Devel- bly have been clarified easily or at opmental Item (NDI) strategy is little cost prior to execution. not at odds with using a methodi- cal and structured approach. How- Because the accuracy and complete- ever, in addition to the time sav- ness of technical data is critical, pro- ings offered by NDI, in this case gram offices need to devote time, senior Army officials outside the money, and effort to researching a po- acquisition chain tried to gain ad- tential supplier’s ability to comply with ditional time by rushing the pro- U.S. TDP requirements. Comprehen- curement cycle. As a result, proper sive reviews of technical data and draw- investigation and confirmation ings are the absolute minimum re- procedures were not used to as- quired. A survey of the potential sure Royal Ordnance’s capability supplier’s manufacturing process and to perform to contract. The pres- configuration control systems are also sure to get the howitzer into U.S. extremely important. production forced the program The real key to success in this area is

45 Acquisition Review Quarterly – Winter 1996

the determination that the production which best satisfies the user’s need in a process in use by the potential supplier timely manner. In order to satisfy that satisfies several requirements. First, the purpose, we have a structured and me- process in use must comply with the thodical approach by which we procure process that the manufacturer says (and items. By using selected strategies, such documents historically) it is using. Sec- as NDI, we can efficiently reduce the ond, the process in use must clearly lag time between requirement identifi- produce the product in conformance cation and need satisfaction. Unfortu- with the applicable drawings. Finally, nately, our acquisition strategies are of- there must be an effective management ten distorted by political realities. system in place Clearly the most difficult problem to to monitor and overcome is that of ‘outside’ meddling Clearly the most document con- in the procurement process. difficult problem to figuration man- Procurement professionals need to overcome is that of agement. be shielded from the unrealistic de- ‘outside’ meddling... The respon- mands imposed by ‘interested’ parties. sibility to en- A solid, logical, and realistic baseline sure that our suppliers are fully aware schedule and process by which the pro- of U.S. Government-peculiar require- gram office gathers information and ments rests on the U.S. procurement makes decisions is critical in prevent- officials involved with the acquisition. ing outside meddling. Program mem- Without a clear understanding of these bers need to be especially attuned to requirements the supplier may very well the political winds and their effects on find that, like Royal Ordnance, it is will- the program. Early identification of ing to comply but it is unable to do so. potential problems are a significant step In the case of the L119, Royal Ord- towards rational solutions. nance was already producing the sys- Early identification of potential tem and their customers were very sat- problem areas is a key to success. As isfied with the results. It was only when with technical data transfer, all aspects the U.S. tried to enforce compliance of contract performance need to be with its TDP standards that Royal Ord- explored early in the process in order nance started to have problems. Early to identify and resolve issues prior to and continuous interface involving rep- award. In that way, potential sticking resentatives from both sides can be an points between the two parties can be effective problem resolution technique. resolved in a cooperative atmosphere, The program office and Royal Ord- rather than in an adversarial conflict re- nance did, in fact, meet repeatedly, but volving around interpretation of a con- it was after the contracts were signed. tract clause after the fact. Had the U.S. By that time, Royal Ordnance was com- conducted a detailed investigation into mitted to standards with which it could the practices employed by Royal Ord- not comply. nance (i.e., technical drawing, configu- The purpose of every acquisition ration management, and documenta- should be to get the piece of equipment tion control procedures) before select-

46 Tech Data Transfer

ing the company as the source for its philosophy was radically different than new howitzers, perhaps the outcome that required for U.S. Government would have been different. If done prior contractors. Additionally, it did not to award, the changes to Royal know that the drawings being used by Ordnance’s process could have been Royal Ordnance did not really reflect made a condition of the award. If Royal the products being produced in any true Ordnance declined to accept ‘our way engineering sense of the word. Finally, of doing business,’ we would be free to because the time was not available to find another source or solution. explore Royal Ordnance’s ability to With the decision to replace the ex- provide an American standard TDP, the isting fleet of M102 howitzers, the U.S. Government did not know until af- Army hoped to procure a major ter the contract was signed that Royal weapon system under the NDI ap- Ordnance would not be able to provide proach. This approach, it was hoped, a Technical Data Package suitable for would get the howitzer into the field use in a U.S. much more quickly than if the weapon production fa- were to be developed from scratch. cility. Thus, we The fault does not lie with Royal Ordnance. However, due to multiple factors, it was entered into a They had proven, over decided that only a portion of the weap- contract with a time, that they could ons would be produced offshore, with party who was produce a quality the remaining weapons being produced unable to com- weapon system. within the U.S. arsenal system. This plete their por- plan, while not unsound, ran into some tion of the agreement. The program difficulty. From the inception of the office was rushed into getting the sys- program, three relatively senior Gen- tem on contract without being given eral Officers applied and maintained adequate time to investigate the full im- pressure on the program office to get pact that the transfer of the Royal Ord- the howitzer fielded quickly. As a re- nance TDP would have on our acquisi- sult, the process, already shortened by tion. The fault does not lie with Royal the removal of development, was Ordnance. They had proven, over time, rushed further. With the selection of that they could produce a quality the Royal Ordnance L119 as the weapon system. However, the method- weapon of choice, events came together ologies and requirements of the U.S. to portend trouble. The combination of acquisition community were totally for- the processes by which Royal Ordnance eign to the decision makers at Royal made the L119, the need for domestic Ordnance. This, coupled with unreal- U.S. production, and the time pressure istic time demands on the decision mak- being applied were directly at odds with ers on the U.S., led to a TDP problem each other. Because time was not avail- which had, and continues to have, a sig- able to investigate and assess the meth- nificant impact on the M119 program. ods Royal Ordnance used to manufac- Although the acquisition of the M119 ture the L119, the program office never has been called a “model of future pro- knew that the British manufacturing curement” (Reece, 1991, pg. 718), it is

47 Acquisition Review Quarterly – Winter 1996

also correct to state that the acquisition sociated with the transfer and manage- of the M119 howitzer should serve as a ment of technical data. ‘how not to’ model for the problems as-

48 Tech Data Transfer

REFERENCES

Armament and Chemical Acquisition Headquarters, U.S. Army Armament, and Logistics Agency. (1994). M119 Munitions and Chemical Command. 105mm Howitzer: Lessons learned on (1985). Non-Development Item (NDI) Non-Development Item Program, Program M119 105mm Howitzer (White Paper), April 20. Author. (White paper), March. Rock Island, IL: Author. Army Magazine. (1986). M119 105mm Towed Howitzer (Equipment descrip- Nathan, J. (1995). Response to request tion), October. for information from Weapon System Manager, M119A1 105mm Howitzer Defense Systems Management Col- System, February 19. lege. (1991). Glossary: Defense Ac- quisition Acronyms and Terms, Sep- Nathan, J. (1995). Response to request tember. Fort Belvoir, VA: Author. for information from Weapon System Manager, M119A1 105mm Howitzer Fahey, K. (1994). Response to request System, January 30. for information from Development Project Officer, M119A1 105mm Nathan, J. (1995). Response to request Howitzer System, December 8. for information from Weapon System Manager, M119A1 105mm Howitzer Fowler, W. (1991). The Royal Ord- System, January 24. nance 105mm L118 Light Gun-An Ugly Duckling Makes Good. Asian Reece, Col. R.G. (1991). The M119 and Defence Journal, January. NDI success story. International De- fense Review, July. Headquarters, U.S. Army Armament, Munitions and Chemical Command. U.S. Army Armament, Munitions and (1992). Briefing Packet, Title: M119A1 Chemical Command. (1987). 105mm Howitzer, U.S. Technical Data AMCCOM FY 87 Acquisition Plan Package, March 17. Rock Island, IL: No. 87-1 for M119 Howitzer, Light, Author. Towed, 105mm, revised October 7. Author. Headquarters, U.S. Army Armament, Munitions and Chemical Command. U.S. Army Armament, Munitions and (1987). Memorandum to LTG Skibbie, Chemical Command. (1985). Deci- Subject: M119 Howitzer Trail Produc- sion Coordinating Paper for Howitzer, tion, March 17. Rock Island, IL: Au- Light, Towed, 105mm: M119, Decem- thor. ber 18. Rock Island, IL: Author.

49 Acquisition Review Quarterly – Winter 1996

U.S. Army Armament Research, De- U.S. Army Logistics Management Col- velopment and Engineering Center lege. (1994). MAM Course Reference (Fire Support Armament Center). Book #1, July. Fort Belvoir, VA: Au- (1987). A systems analysis and ratio- thor. nale for the selection of the UK L119 howitzer for the Light Infantry Division Artillery. April. Picatinny Arsenal, NJ: Author.

50