Should the Navy Design its Own Ships?

This html article is produced from an uncorrected text file through optical character recognition. Prior to 1940 articles all text has been corrected, but from 1940 to the present most still remain uncorrected.  Artifacts of the scans are misspellings, out-of-context footnotes and sidebars, and other inconsistencies.  Adjacent to each text file is a PDF of the article, which accurately and fully conveys the content as it appeared in the issue.  The uncorrected text files have been included to enhance the searchability of our content, on our site and in search engines, for our membership, the research community and media organizations. We are working now to provide clean text files for the entire collection.

ments, was performed by the Navy. Its results were ¹¹¹ passed on to the competing contractors. At the of the next, or contract definition phase, the requirement for a new ship design was transmitted ¹¹ the prospective contractors in terms of ranges of ?^e<

formance characteristics. Their responses were to re:

a balanced design solution considering performance teria, design standards, and production techniques.

Preceding pages: Anchors and chains for the new Tarawa are laid out alongside the ship at the pier where she is being completed by Ingalls Shipbuilding Division of Litton Industries, Pascagoula, Mississippi. The design of ships begins with the appreciation by operators or politicians of a need or an opportunity, and proceeds through several stages to detail design where such matters as the number of anchors, their type and size, how much chain, and its type and size, must be decided. A grand conception may be ruined by poor execution of detail; good detail work cannot save a poorly conceived ship.

In this paper I am going to examine the causes for the recent return to the Navy of the design of its own ships, and to explain how this change resulted partly from disenchantment with the Defense Department’s previous method, partly from the Department’s inten­tion of getting control of the rising cost of acquiring new weapon systems and partly from its recognition that naval ships are significantly different from other weapon systems and therefore require a different ap­proach to their design and acquisition.

First, the events and objectives that led to the shift of ship design from "in-house” execution to private industry about ten years ago will be reviewed. Next, I will discuss the problems which were encountered as a result of the new approach, followed by a discussion of the environment of the early 70s and its particular constraints. Before presenting alternative options for accomplishing the design of naval ships, the peculiari­ties of such ships as they relate to their design ap­proach will be described.

As a result of this detailed discussion, it should be clear that the preferred way is that in which the Navy has in-house design control up to the detail design phase, that is, one in which all the major design deci­sions are made by the Navy.

The return to in-house execution a few years ago revealed some new problems and highlighted some old ones, all of which I will identify and discuss, along with some of the changes that have been made by the Navy in order to overcome them.

The Objectives of "Total Package Procurement"               ^

When Robert McNamara was the Secretary of D^c fense, he initiated the so- called "Total Package Procuf^£ ment” concept for major weapon systems, includi*¹? naval ships. This entailed a process called Conccp Formulation/Contract Definition (CF/CD) whi^¹ through competition in industry, led to the develop, ment of a design and the contract for procurement⁰ new ships built to that design. This process was sp⁰¹¹ sored by the Department of Defense.

Three ship acquisition programs were defined aⁿ developed by the CF/CD process: the FDL, the LH-^V and the DD 963. The FDL program was cancelled a‘^tCj the completion of CF/CD, but before the award ⁰ a detail development and construction contract. ^ other two contracts went forward and their prog^re' has been well publicized.  j

The Total Package Procurement (TPP) concept a" the CF/CD process were not necessarily insepara^ one could have been employed without the 00 However, in the case of the three ship acquisid^ projects of that era, the two techniques were combin'- "Concept Formulation,” which was intended to ^v£l ify the compatibility of all the performance requ’^rC

0

d>I flii*

process emphasized the industry’s responsibility for' ultimate satisfactory performance of the ship. Navy’s role during this phase was to monitor and g^ul the builder, should his endeavor take a turn wfo experience had shown to be disadvantageous.

The end product of contract definition was a P^r</ posal from each would-be contractor for Total Pac**»j Procurement. This included a set of ship plans ^aP specifications developed by the contractor in resp°T to the Navy’s requirements, together with deta‘ management, facility, logistic support, and other which were to be followed during the detail devel°t ment and production phase. As a result of a leng^tP, source selection process, the Navy then chose ^ contractor whose ship design and plans for construct‘d were considered the most cost-effective. This requ‘^f ₍ an intensive effort over a period of several months a*¹ receipt of the proposals, and led finally to the neg<^‘ tion of a single contract with the successful firm- ¹ _f contracts that resulted were for the entire product'¹ run of each ship class.

w i/^r'ⁿ^^e ^^ene^^c this approach, it was hoped, tr ^,U j     ^rn°dernization of the shipbuilding indus-

plica ^eS'^^{n ant}^ P^r°duction processes through the ap- ^Catlon 'he latest techniques of management and th^^IJt'^ons analysis. Previously, shipbuilders dealt with based ^construc''^{on as}pect only. Their bidding was 'he M^{CntIrdy on a} design developed and provided by _re • ^3Vy' ^he industry had no ability to take just _a ^Cl^Ulrernents and, from these, design a ship. There were ^op h^fms which had some preliminary naval _anj ^es'8ⁿ capability, but only one of these, Gibbs ^ox, actually had prepared preliminary designs for _mu-naval ships. To be able to do this, a company ‘"clin ^    P^eoP^^{e w}b° hy training and

^ent -?^at*^{0n are} able to do that part of the design which dev |^{S Startin}S ^w''h ^{a clean} sheet of paper and on it ^s'ated°P ^dternat'^ve concepts, all of which meet the build P^er^^ormance requirements. Similarly, only five des' ^ ^ ^not only experience in Navy contract shi'^ⁿ ^^{Ut dso} P^ossessed yards capable of building _ne^^S °f 'he FDL and LHA size. Obviously, with the had ^appr°^{ach t0} procurement, the successful bidder _ncer.^{t0 c}°nsider joining analytical, management, engi­rt and production talents into completely inte­nd teams.

(j)          'he objectives of the new procedures were:

^vati ° *^nPuse 'ⁿ'° 'he naval ship design process inno- ^Sefim^{eileSS n0t} °ⁿ^^{y prom} shipbuilding, but from other bilit ^{CntS op}P^r*^vate industry; (2) to introduce produci­ble ^ ^lnt° ^early ^sign decisions; (3) to define precisely t_0r ^resP°^ns'bilities of the government and the contrac- fi_x^^{t0 prov}'^^{e a} good basis for firm fixed price or ''ed ^^nCe ‘^ncent'^ve contracts (given that CF/CD is rexjuJ^{0 TPP}^’ W ^t0 establish realistic performance '"du ^ments (^chis assumes that the engineer in private tvhat ^ ^^nows better than an engineer in the Navy and _e *^S ^^aS^^e)’ ^an<^ (⁵) ^t0 strengthen the industry In ₍i'^U'^p *''° be responsive to future naval programs. ha_s ^ ^Ore^⁰*ⁿS’ ^a design for which "producibility” ^Ur>de ^CCn ^'^ven consideration means that the designer p^r°_Ce^^tand^s and has thought through the production 'o _e ^S0 .^tbat the shipbuilders will not have to resort built ^pCns‘^ve production processes to get the ships

f^or the Failure of '^-^Jjfkage Procurement Tip

pt_Q^l^1S approach failed to solve such fundamental fo_rm^{Cms op} major weapon acquisition as meeting per- F_0r ^2riCe objectives within projected cost and schedule. F.jjj^'doce, one need look no further than the C5A, ’ and DD 963 programs, to name a few.

While at first glance the objectives seem reasonable, a close examination reveals that some of them are related purely to the method of procurement and have nothing to do with the method of design. Others are related to accomplishing the design, but they are based on the erroneous assumption that CF/CD is the best way of achieving these objectives.

Let us examine, one by one, the five objectives of this approach. It will be clear that they either were not met by the execution of CF/CD or that the objec­tive itself was based on erroneous assumptions.

(1)  The design phase for a new ship class to be introduced into the fleet is not the time for the dis­covery and implementation of untried technology. Rather, it is a period for integration of existing tech­nology. Because of the risk, major technological in­novations should not be accepted in an acquisition program for the fleet. Therefore, the argument for the best technical innovation through the competition of private enterprise is not applicable. In the space pro­gram, technological breakthroughs were necessary and cost was almost no object. But that is not so for conventional surface ship programs. In such programs major innovation is seldom expected; and it is welcomed only at minimal risk and if it promises lower cost. A stable work force that has developed lines of communi­cation to solve the problems of integration is far more important than the inventiveness which could be achieved by involving the large resources of industry.

Of the three to three-and-a-half years that elapsed between the time the Navy started the CF/CD process and a contract was signed to build the ship, less than one year was devoted to designing the product. There­fore, even if innovation should have been an objective, there wasn’t enough time devoted to it. As a matter of fact, much of the effort expended by the hotly competing CD teams was directed toward "selling” rather than toward design development.

(2) The second objective assumed that it is possible to introduce producibility at the early stages of design. My experience has shown that, since the design is quite fluid up to about the middle of the contract design period, it is difficult to introduce such considerations before that time. Of course, at that time it is possible to call in representative shipyards and let them comb through the Navy’s guidance drawings and specifica­tions and ask them to suggest improvements.

(3) While the definition of responsibilities is desira­ble, the CF/CD process did not guarantee it any better than did other possible methods, as I will show later. In fact, there are good reasons why a better chance exists to achieve this objective without involving the producer in the early stages of design when the Navy hasn’t yet decided exactly what it wants. It is easier

to define the responsibilities of the parties when they both know what they are talking about; however, this has little to do with who actually controls the design as long as both parties to the contract have a reasonable time in which to examine the technical package and have a chance to enter their comments into it before signing the contract. In any event, this objective applies only to the lead ship, because later ships built to the same design benefit from a complete detail design and that is always done by private industry.

(4)The strong emphasis on low life-cycle cost with­out good quantitative means of specifying effectiveness

tends to result in a ship design meeting only the s<j called "minimum requirements” in the ways of sp#⁴ endurance, reliability, quietness, and so on. Howcv* the requirement document itself can define only imp⁶'¹ fectly the minimum effectiveness for a system as cot¹ plex as is one of our new ships. Therefore, the result!¹¹!¹ design could fall short of meeting even the minirn^ul11 requirements. Thus, it requires quick and easy m^¹¹ of dialogue between the requirement definer and tl* ship designer; this is difficult through a contractu¹ barrier.

Also, during contract definition, only negative g^ul“

ary demand for engineers.

Also, no single contractor could justify the m³¹^, tenance of a large and competent design team with⁰ a constant design workload.                                                          . ^

The outcome of the three competitive ship de^slr

ance could be provided by the Navy because of ^[1 competitive nature of the process. As such, the pr°fj, did not enable any designer to obtain the best avail³ information from the Navy’s vast storehouse of &'

(5)                                                                                                                                                                             Competition among several companies f°^r large shipbuilding CF/CD program placed a very h&j but short-term demand on the small community ship designers, resulting in an unreasonably infla^t|0

^e 963

Th,

* LHA

Th,

Contract The fdl

contracts that were accomplished through the CF/CD P^r°cess was as follows:

Competitors

Litton—winner Lockheed General Dynamics

Litton—winner General Dynamics Newport News

Litton—winner General Dynamics Bath Iron Works

Ai^{tCr eac}h competition there was a gap of a few ^an I ^{S Dudn}g ^t^^lat P^eri°^d every competitor other . 'tton disbanded its design team. This was because ^and rl' °° ^exP^enst^{ye t0} maintain a team of the caliber Sun ‘^VerSity voiced for CF/CD without government petg °^rt’ °^r ^^ecause the company decided not to com- the i,^{n anot}^^er LF/CD program. In the case of Litton, n_ot ^ree-time winner, the design team, even though ₃p ^lsbanded between programs, had largely dis- ,j_CS]^Cared hy the end of 1973, because the ships it had P^rog^rcssed into production, when only ^0r changes become necessary.

Early Seventies

tUtJ¹ Edition to the problems with design and pro- CP/_c^^ent U. S. naval ships using the TPP and iv ^aPproach and the changing of the Defense carl ^artment’^s administration in 1969, the late 60s and j_n . were also a period of tremendous expansion _SUant^C ^^0viet surface fleet, notable not only for the d_esj 'T hut also for the quality and diversity of the fJ^{S rc}P^rescnted- 1° order to counteract this threat, Which ^avy undertook a new ship design program of _s^. ^resulted in the largest number of different classes \_x ‘P^s undertaken concurrently since World War II. put ^{at same} time, the Department of Defense was cam_t^{UnCiCT cons}*derable financial pressure. Congress be- in ]| ^c°^ncerned over the size of the defense budget goy ° ^{C otdler} social needs. Pressures to limit total the ^Crnment spending and a general preoccupation with a ^ec°ⁿ°my, the environment, and energy made DoD ts_Cai^{1TlC tar}S^et for budget cuts. Well-publicized cost sur_e ^atl°^ns of major weapon systems led to more pres- iw r.ⁿ ^tmgress and the Executive Branch to initiate _As^hscal controls.

^{a res}ult, three major studies were begun in 1969-

All of them addressed the problems which plagued major weapon acquisition programs, and one, the "SON Pricing and Cost Control Study” performed under the direction of the Chief of Naval Material, specifically addressed new ship construction.

In response to all of these events and the recom­mendations of these studies the Department of De­fense, under the leadership of Deputy Secretary of De­fense David Packard, began a number of actions aimed at controlling costs of new weapon systems by means of controlling the technical and financial risks.

The policies have been summed up succinctly in a report by the Senate Armed Services Committee:

a.   reducing concurrency,

b.   designing to cost requirements,

c.    using prototypes,

d.   requiring hardware competition,

e.  reducing radically the size of industry design teams,

f.   increasing independent OT&E prior to procure­ment decision.

Of these new policies, items a, c, d, and f have been largely consolidated under the well known "Fly Before Buy” concept and item b has become known as "Design to Cost.”

In 1973, Dr. John Foster, then Director of Defense Research and Engineering, defined Design to Cost as "trading performance for cost until we are assured that a balance is achievable wherein needed military per­formance can be provided at a price we can afford for the quantity we need.”

Peculiarities of the Naval Ship

To evaluate the worth of alternative approaches for the design of naval ships so as to choose the best, the design phases, along with the peculiarities of naval ships as related to their design, need to be understood. These peculiarities require a unique solution to the problems of design and acquisition for naval ships when com­pared to the design and acquisition of other major defense systems, such as airplanes, tanks, and missiles. Naval ships are unusual because of their large size, their nonhomogeneity, and their complexity.

These qualities dictate a very detailed technical defini­tion early in the design stage, requiring greater sophis­tication than do other defense systems in cost estimat­ing, in the application of margins, and in contracting provisions which avoid unreasonable risk to the con­tractor while, at the same time, offering incentives to control costs. The large number of different engineering disciplines required to design the ship, and the intricate

to perform the various studies leading to requirenv determination led to the creation of the Center

efit*

for

re'

iSt*

interrelationships involved, strongly influence the organ­izational structure of both the requirement definers and the ship designers.

The ship is a weapon-carrying vehicle, and the tech­nological disparity between the platform and the pay­load has ramifications on the design process. The fact that the platform, with its fairly weak performance link to the weapon payload, is a conglomeration of tried and proven (nondevelopmental) systems means that ship design is basically a period of integration and not of invention, which should influence the environment within which the design is accomplished and the time of contracting; that is, whether we want to contract for the design and production in one package (CF/CD) or in separate packages.

Another peculiarity of the naval ship is the long period of its development (many years) and the fact that we buy ships in small numbers and at a high unit cost. The budget scrutiny attracted by high unit cost results in the need for dedicated project management attention and responsiveness to inquiries from Congress on down. Cost estimating is relieved somewhat by not having to cope with the uncertainty of predicting sig­nificant learning curve rates but, on the other hand, it never has the benefit of a completed prototype on which to estimate costs of follow-on ships. Contracting must find means other than prototyping to protect the Navy’s interests, since ships procured in small numbers are hardly conducive to prototyping.

The expected outcome of design of any defense system coupled with the just enumerated peculiarities of naval ships dictate the following objectives relative to achieving a good design definition: (1) to develop a design which is capable of achieving the desired performance, (2) to be able to estimate in advance of the contract for production what the cost per ship will be, so realistic trade-offs can be made within the con- traints of a level defense budget; and (3) to assure with a high degree of confidence that the expected perform­ance and estimated cost will be met.

Options for Accomplishing the Design of Naval Ships

The time needed to design naval ships is shown in Figure 1, along with the generic stages of the design. (Times vary, of course. The PF took about three years. A CVN might take over four years to design.) The various participants, such as CNO—the user; CNM— the user’s procurement agent; and industry—the pro­ducer, are also identified. Naturally, the Navy could establish a contract with industry at any stage or lump together as many phases as desirable. For example, in contrast to the total package procurement, when all

phases of design and production were in industry* hands, the Navy could go to the other extreme aⁿ produce the lead ship in a naval shipyard and ne^vet need a contract for the design or construction of tht ship.

There is no qeustion about the fact that the can’t ask a contractor to determine its requirement, an therefore this phase has always been performed by tl* Navy. Of course, the job is large and contractor supp^0(t is needed for specific studies in the way of threat evab ation, mission analysis, and so on. This need for peop¹

Naval Analysis (CNA), a non-profit company working exclusively for the Navy. However, the task is so la^rf> that, even though CNA exists, there is still a g^(Ci'‘ deal of contracting to private firms for specific studio

Up till 1966 the next two phases, the design con identification and the design of the system to a l^eV of detail adequate to contract for production of ^ ships, were accomplished by civilians and a few offi^cet in the Bureau of Ships. Beginning in 1966, the T with CF/CD was the method used in designing *ⁿ procuring naval ships. The question that faced Navy at the beginning of 1971, in view of a D° directive prohibiting TPP with the CF/CD appro*⁰ and in view of the problems with respect to its apP¹ cation to naval ships, was how and where to accomp'¹* the necessary designs.

The early stage of the design process is primary one of engineering trade-offs. The designer develop several "ships,” each reflecting a different set of op⁰** tional requirements (which in turn are the result ⁰ a wide variety of perceived needs and pressures). Th^e:’ are translated into comparable ship acquisition ^c° estimates. These alternatives, each satisfying a unifl¹ set of operational requirements with an associated cO*' are then traded-off until the decision maker OpNav sponsor) selects a combination which offers desired balance between operational requirements *^p cost.

In his quest for reasonable cost, the OpNav sport⁵' may have to consider matters besides the trade-off mission requirements, involving austerity and innO^v' tive features such as whether there will be one or ^ missile launchers, whether there will be one helicop¹ or two, or to what degree the propulsion plant be automated. He can then decide which of these all⁰ him to achieve his cost goal.

Therefore, in the early stage of design it is import* that those involved in such actions as the repeti^tl process of balancing requirements against implied enjoy freedom of dialogue between the requirerne⁰. definer and the ship designer. This is extremely &

Figure 1 Alternative Methods for the Development of Naval Ship Design (The arrows and dotted lines show continued interactions.)

O O

O

PHYSICAL

DEFINITION

FUNCTIONAL

DEFINITION

CONCEPT

SELECTION

PRE CF/CD (PRE-1966)

CNO

REQUIREMENT

DETERMINATION

TIME SCALE (AVERAGE VALUES)

MILESTONE:

CNM

CONCEPT DESIGN

PRELIM.

DES.

CONTRACT

DES.

CONCEPT FORMULATION CONTRACT DEFINITION (CF/CD)

PLUS

POST CF/CD (1971-CURRENT)

(INCLUDING TOP LEVEL REQUIREMENTS/ TOP LEVEL SPECIFICATIONS CONCEPT)

CNO

CNM

REQUIREMENT

DETERMINATION

lATIOIj

■ o n

CONCEPT DESIGN

X

X

PRELIM.

DES.

CONTRACT| DES.

TOTAL PACKAGE PROCUREMENT

DETAIL DESIGN AND PRODUCTION

From Design -to-Cost of Naval Ships'’ by Reuvcn Leopold, Otto P. forts and John T. Drewry. Presented for the 1974 SN±j\S1E annual meeting and included herein by the aforementioned society.

adequately defined design is one in which later

on

&

cult—if not impossible—through a contractual barrier.

This phase is extremely important since "Require­ment to Cost” (i.e., the early trade-offs of requirement: one missile launcher vs. two per ship, a large sonar vs. a smaller and consequently less costly one, or a very large magazine with many missile reloads vs. a smaller one with few missiles) has an impact on the cost of the ship many times more than that which can be accomplished during the "Design to Cost” phase, in spite of the fact that significant achievements are still possible through design trade-offs once system require­ments have been frozen. During these phases of design, a stable organization where people are used to com­municating with each other is more important than the number of "inventive” persons involved.

Another important aspect of the early stage of design is the resulting cost estimate with which the Navy goes forward to DoD and Congress and which then is used for budgetary purposes.

This product is also strongly dependent on the "adequacy of definition” of the design since the design is supposed to be the basis for the cost estimate. An tl>⁽|

detail designers will not discover situations requid’f important redesign before they can get on with d¹ work.

In a recent examination of ship costs, it was covered that in 17 of the 35 cases studied, inadeq^u; estimates contributed to the "overruns” (see Table A recent Design-to-Cost study reported that:   i

"A set of 32 ship case histories were analyzed it was found that the aggregate, latest estimated ^c°^ of the ten ships that had budget figures based Class C estimates was 10 per cent under the aggreg³p budget; whereas, that for the 22 ships with Class estimates exceeded the aggregate budget figure by per cent. (See Table 2.) A Class C estimate is deri'^f from a substantially more detailed definition of a d¹ so the conclusion that better definition is needed >^l lows—both logically and from the evidence.”

Thus, "definition prior to acquisition” is the p^r'^f1’ concept that should determine the method of design¹¹ naval ships up to the detail design phase.

However, no contractor can afford to maintain

Three stages in the life of a new ship. At far left a sketch (this one actually made about a year ago) shows what the designer has in mind; in the center four ships in various stages of construction at the shipyard. From left the 1)1) 963 and, successively, the 964, the 965, and (barely begun), the 966. The photo below shows the Spruance, name ship of the new destroyer class, on her trials in February. In the bottom view we see the SQS-26 sonar which has so much influence on the shape of the bull and on what other sensors and weapons are carried by the ship.

Wg_c and

^w°^rklo ,i ^ComP^etent design team without a constant ^;ign

^vate            periods (years) to sustain even three pri-

government.

design ^natUr‘d answer appeared to be some sort o ind_u, P^artnership between government and private

c

^Khip

_J                                     X       ---- -------- C>       ------

de_s; ^at ' ^spite of the current upsurge in naval ship ^act*vities, there are not enough programs foi

^singl_e°^r^^an’^Zataons *ⁿ competition. This points to £ of ^cer,ter for naval ship design. But since our style t°nm^Vernment doesn’t allow the subsidizing of one so|_Ut- ⁿ' .^t0 maintain its design capability, the onl) in _tp ^ ^ts th^e consolidation of the design activities

Earl          avy itself.

1971 the question that needed answering f>oli_c;_e “^at organizational and contractual method and exp . ^can maintain a large, constant, competent, fo_rce^'^nced) cost-conscious, and unbiased design work ^nsidg ^cl^uesd°ⁿ had to be examined with due of _tj_le £^atl°n for the growing desire to reduce the size T'he ^e<aCra^ government.

‘gn |Ustry'

ⁿext question was "how are we to define thi: ⁵” While the detailed relationship whief

perfori^ staff \

1.             Planning to be done by the Navy.

2.             Major decisions to be left in the Navy’s hafl

3.Parts of the design work to be done by the N*¹'' (to maintain expertise in such work).

4.              The bulk of the design work to be

by independent design agents (not on the shipbuilder and thus free of hardware conflicts f interest).       I

5.              The design agents to be located near the NI

designers so there can be daily face-to-face comm^ul1' cation with the Navy.      I

6.The Navy’s planning to provide for stability 1 the design agents’ work force.

In order to achieve these goals, the Naval Engineering Center was required to increase its ^caP^f,) to handle the large volume of new ship designs. ¹ ^ 1 was done by shifting some of its work force            !

in fleet maintenance support to  new design work,       j

ing their old tasks to outside concerns. Simultaneous ^ incentives were provided for the establishment ⁰ ^ stable, competitive community of design agents lo# in the vicinity of NavSEC which could be tasked ⁰⁽¹j quick turn-around basis through proper contraC¹¹ j channels.                                            _Q[ |

Today NavSEC is responsible for the execution all U. S. naval ship designs. It has dedicated the eq^ul\ lent of 400 men per day on the average to new ^s i design work along with a private design agent force of about 200 more people, most of whom located near NavSEC. Thus, in the last seven years, ₍ have circled back to in-house design in order to ach‘^e the goals delineated earlier.

Naturally, in the transition to a new mode of op^e

+

TOTALS

| ^cr\ j j NO

mm

1 S U. r- ^¹

du|5L|3

■— w :  O j O

^ i X I X j T j X I

o

cy

h-* c*-

u-, ^

< 

GENERAL

PROBLEM

AREAS

ASPECTS OF PROBLEM AREAS FOUND IN CASE HISTORIES

11

Developmental

Work

Building experimental ships and funding developmental components under SCN and cost

constraints

Planning

and

Programming

Conflicts with PPBS cycle timing, lack of formal acqui­sition plans, inadequate ship definition, naval/private asvard effect, program instability

17

Estimating

Estimating inaccuracies, 'ball park' estimates, lack of ade­quate data base

28

Economic

Conditions

Escalation, profit factors, influence of competition, builder’s workload, delayed award effect

13

Pricing

Practices

Directed changes in estimates, statistical factors, funding policy

24

SHAPM and Management System Controls

Fragmented authority and con­trol, inadequate financial and information reporting systems, manpower shortages in Headquarters and SUPSHIP

24

Disciplined

Decision

Making

Changes made, work begun, directives issued without knowledge of cost impact, changing requirements

Conflict, ambiguity, omission in specs, late GFM and GFI impact, change orders, claims

Naval                        Priority system, overtime Urn-

Shipyards           I itations, delay/disruption,

personnel ceilings, material acquisition

Brief Reason for Difference

Ratio of Aug ’68 Estimate over Estimate Classification*           Submitted to Congress'

Inadequate cost data base for REHAB Pricing Action and Char. Change Inadequate Cost Data Base Favorable Market Conditions for AFS3-5 Pricing Actions, Specification Change and Characteristic Change Developmental Nature of Project 2 Contractors Involved, Strike and Flood Damage

Research and Development Project Overestimated Projection of Escalation Poor Scope Definition, Characteristic Change, Delay and Disruption Poor Scope Definition, Delay and Disruption and Inaccurate Estimating None

Poor Market Conditions and Specificati°ⁿ Improvements

Inadequate Design Scope Definition and Inaccurate Estimating Inadequate Design Scope Definition and Inaccurate Estimating Poor Design Planning, Characteristic Change, Delay "in Shipyard Characteristics Change, Delay and Disruption

Insufficient Data Base and Good Market Conditions Contractor Claims Growth Factors Overestimated Surface Missile System Improvement Program

Poor Definition of Scope Poor Definition of Scope Inadequate Design, Poor Market Conditions

Pricing Action and Change in Market Conditions and Economic Factors D^u( Deferral Period

Pricing Action, Major Change in

Characteristics, Closedown and Chang^e in Building Yard

Overestimated Projection of Escalation ^ Good Market Conditions and Overestin¹* of Growth Factors

Equipments of Development Nature.

Requirements, Late Equipment DeH^ve Overestimating of GFM and Growth Factors

Major Change in Characteristics After Contract Award and SUBSAFE Major Change in Characteristics After Contract Award and SUBSAFE

our

'A

S^'P^S, four experiences with cost estimates. Top view, the carrier John F. Kennedy ^)» f^or which the cost estimate was right on target. At right, the carrier Midway t_u,j^ f¹) after her latest modernization. The cost of that modernization was more than

*hi

C‘ ^t^^>e price estimated. At left, two ships whose building costs were less than anticipated,

combat store ship Sylvania (AFS 2) and the ballistic missile submarine Will Rogers

659).

of

was that BuShips reported directly to the Secretary

of

ment between the engineer in the design branches

of ships in the Navy, found the highest level orgs”. ization (the staff to CNO) deficient in a number ⁰

ways:

requirement derivation was to be achieved, thus

id>

tion, especially under the accelerated change experi­enced today in government, in technology, and in society at large, it was found that reverting to in-house design did not mean "doing our thing as we did it before 1965.” There were a number of critical problems which arose from circumstances that didn’t exist before 1965, or that were not then recognized as problems.

I will examine two such problems which were signifi­cant, and the solution the Navy formulated to over­come them.

Two Problems and the Navy’s Solutions for Them

The Navy’s need for a new class of ships is identified by OpNav in one of several ways:

1. By Op03 for surface ships and by Op02 for sub­marines simply as a result of (a) the need to replace old ships or (b) a major new shipborne weapon system is ready to go to sea and it would be more costly to convert an existing ship type than to design a new one.

2.By Op96, as a result of having examined new threats, translated them to new missions and scenarios, and possibly having decided that new hardware is needed.

3.         By CNO himself, as a result of advice by the CNO Executive Board, other admirals, or anyone else who can influence his thinking.

Regardless of the source of the requirement for a new ship, it needs to be translated into a specific hard­ware definition. The old method of generating and agreeing on requirements, together with the second step of translating such requirements into hardware, such as boilers, pumps, and generators, was found to be inadequate with the return to in-house design in 1971.

The inadequacies were related to both organizational relationships within OpNav and the Material Com­mand (specifically NavSEC) and to the lack of a ra­tional, traceable, and well documented dialogue be­tween OpNav and the Material Command. Let us first discuss the organization issue.

Changes in the Navy’s Organizational Structure for Ship Design

Figure 2a depicts the ship design organizational structure of the Navy which remained almost totally unchanged from 1939 to 1966. This organization was characterized by seven bureaus participating in the design process. On the CNO side of the house, the Ships Characteristics Board (SCB) specified only re­quirements. The design function was concentrated in BuShips Code 410, and it was there that the ship design developed with participation from the other bureaus and CNO. A noticeable feature of that organization

the Navy. There were three principal layers of manag^e'

BuShips and the Secretary of the Navy—the final deci­sion maker and the authority for funding.

By contrast, the ship design organization of CF/CD and post CF/CD era is complex (see Figure* 2b, 2c, and 2d). Certainly there are more participant and several more layers of management. The ship d^£ sign function is no longer carried out in NavSea, m^{ current successor to BuShips, but rather in a separate command, the Naval Ship Engineering Center (NavSEC). The latter reports to NavSea and is locate ten miles away. Also, there are two layers of manage­ment, NavMat and OpNav, between NavSea and the Secretary of the Navy. The major system comp⁰' nents—the ship, weapons, and electronics systems^ became not only the responsibility of separate cotf' mands but also of numerous project managers with”¹ these commands.

However, the ship design process must always ⁰ closely aligned with the ship acquisition metho • Therefore, in concert with the 1971 changes in D°P policies, the organizations involved in ship design, bo< in requirements derivation and in design execution needed to reexamine their organizational structures an make the necessary changes.

The Change in the OpNav Organization. The chang^cj to DoD policy in 1971, which dictated stronger contt° by both the Navy and DoD over the design evoluti⁰”

There was no single organization for the establish ment of ship characteristics for all types of ships. effort to establish the characteristics of a new ship f always lengthened by interminable delays, extending into months, while the various concerned offices ^0,1 the CNO’s staff debated the specific requirements >°

the ship.                                                                  . ₍

There was no standard procedure as to how ¹

UV11 » t*w*_v** "                                                                          ~ ^w —--------------------------------------------------------- 7   .

new program that came along was left to its o* devices as to how to do it.

Since the process was not often done the same " '_ twice, those who were to establish the requirem⁰¹¹^ were almost always unfamiliar with what they ^ to do.    _c

By shifting the lead from office to office, no ⁰ organization had the breadth of competence to han the task effectively.

(2d)

1974-PRESENT

SECNAV

USERS

CNO

DEVELOPERS

—V

T

SHIP ACO. PANEL

CNM

Ar^

(DIALOGUE) j

\JtG!E.UTy. '•’SHIP ACQUISITION AND IMPROVEMENT

_____________________ OVXO                 \OM AA_ RELA.TVONSH1P

lead

Ao

Smsition and Improvement (Op-97) within OpNav (see Fig_ure 2c).

Op-97 was made responsible for the intensive man- anient of the requirements and their compatibility the evolving design that were needed to cope with increasingly severe fiscal constraints, the economic ^CSC:>lation, and the rapidly expanding technology asso­rted with the military threats. Soon thereafter, the 'tt’ctor of Op-97 was elevated to the rank of vice ^admiral.

AH this set the stage for the selection of participants ^{W 0} could be spared from the organization rather than ⁰ those who were best qualified for the job. This ^naturally had a negative impact on the end product, ^Such ^fhat scenarios were inconsistent with the available ^eS^uipment or technology; these in turn could be obsta- ^es to approval by reviewing authorities in DoD. In 8^eneral, it also led to inefficient use of people.

¹ was essential to correct this situation at the time ^en 'design to cost” and "fly before buy” were gain­ing momentum throughout DoD. As a result, a num- _L^er of studies were conducted in 1971 and early 1972, ^lng to the establishment of the Office of Ship

On a corresponding schedule, other offices in OpNav ^lch were in the business of selecting ship charac- q ^stlcs relinquished that function. For example, P'36, the Ship Characteristics Board, disappeared.

’tionally, those in Op-05 dealing with selection Along new air systems were shifted to, or had addi- °ⁿal duties in Op-97; the same went for parts of the Op-02, for submarines.

^efi as before were the basic organizations of the *0. (Op-02, -03, and -05) as sponsors, Op-098 as ^ budget manager, and Op-096 as systems analysis, ollowing the retirement of Admiral Zumwalt and ^ ^e f^^a§ °(b^cer reassignments within OpNav, Op-97 ^as disestablished in August 1974, after two years of cessfully leading the requirement/design dialogue. ^ee Figure 2d.)

.. ^^0rtunately, however, the concept of Op-97 didn’t ^lsappear, thanks to the establishment of a Ship Ac­q ^lsitl°n Panel headed by the previous director of P'97. I hope that this latest reorganization will not A m loss of the advantages formerly gained.

\^anizational Changes at NavSEC. The reorganiza- ^ln 1966 which created the Naval Ship Engineering ^nter combined the function of fleet support (that is, ^^a*ritenance) with that of design. The hope of the ten ^ ^WaS ^^{at r}^^{e ex}P^eri^ence with operational main- pjnce problems would be fed back into the design 1J** *^{n or}der to improve design. However, between ^ . ^and 1971, during the CF/CD era, there was little ^Csi§ⁿ activity at NavSEC. Thus, the NavSEC opera­tion tended to become predominantly one of fleet support. As a result, NavSEC lost a number of compe­tent ship designers, and their loss was amplified by the accelerated retirement of many who joined the Bureau of Ships just before World War II when we were in another upsurge in naval ship design.

During the period of 1966-1971, NavSEC’s role in ship design was largely that of review, comment, and approval (or disapproval) of privately-developed con­tract designs, along with involvement in concept for­mulations, rather than in the actual execution of de­signs.

When NavSEC reentered the field of ship design in 1971, the traditional concept of subsystem excellence (if each organizational unit designed the "best” possible subsystem for a ship, the "best” total ship would result when these "best” subsystems were combined) was forced to accommodate itself to the concept of com­promising subsystem performance for the sake of total system performance and possible system cost reduction. It was absolutely essential to consider the ship as an entity throughout the design process, with decisions on various subsystems based on what is best for the ship as a whole. The problem basically became one of attaining control of the ship design and development process to provide design excellence within the new cost constraints on both the end product (the ship) and the process (the design) itself. NavSEC has acted to meet this challenge by making substantial changes in its in-house design processes. These changes were intended to provide the top-level visibility and control necessary to manage today’s severely budget-constrained designs with accelerated schedules.

The salient features of the design process resulting from these changes are formal planning, full documen­tation, strong centralized control of the process, and total early involvement by all functional divisions in the effort to fit the constraints of space, weight, and money. Thus OpNav is given detailed information on how the design money will be spent, concrete assurance of value received, and assured continuity of the in-house design process. Within this new method­ology, NavSEC has to assume strong internal control and positive realignment of its organization.

NavSEC has formally assigned "total ship respon­sibility” to SEC 6110, newly retitled as the Ship Design Division (rather than the Concept Design Division). In its new role as primus inter pares, NavSEC 6110 integrates and coordinates all design phases from the first feasibility studies through the preparation and completion of a bidable package of drawings and specifi­cations. To carry out these efforts, a design project team is formed within 6110 to manage each specific program. These teams are composed of members from NavSEC

6101, 6102, 6105, 6110, 6120, 6140, and 6170, other parts of the Naval Sea Systems Command, and other systems commands. These teams are responsible for bringing together the expertise and capabilities of all functional divisions and offices within NavSEC in order to produce a balanced and coordinated design. This has, in effect, combined the three organizations which, in a somewhat overlapping fashion, were previously re­sponsible for various phases of the design; that is, the design management office (6103), the preliminary de­sign division (6110), and the contract design division (6120). Through these changes, NavSEC has reasserted its responsibility for the entire design, from original concept through the contract design. The design man­ager has the formal authority regarding what is to be done on a project and when it is to be done, while the functional managers concerned with such disci­plines as "lube oil,” "hull structure,” or "pumps,” determine how the support will be given.

The Patrol Frigate (PF) was the first ship design begun under this new form of organization and it was completed in April 1973. The design for the Sea Con­trol Ship was completed in December 1973 and the design for the CSGN and CVX are about to enter the pre­liminary stage. Standard procedures for design man­agement were developed during the course of these efforts, and are being refined into a guide for ship design.

Communication between the Requirement Definer and the Ship Designer     

In addition to the required changes in OpNav and NavSEC, the means of communication between the two needed a thorough reexamination and change.

Specifications constitute the vital technical part of a contract between the user and the producer of the product under consideration. In our work, specifica­tions have primarily been associated with the contract for construction of the ship; the parties involved are the Navy on the one hand, and a shipyard (usually private), on the other.

It is often overlooked that the relationship between the Chief of Naval Operations (CNO) and the Chief of Naval Material (CNM) approaches that of a cus­tomer to a contractor. The responsibility of the former is to state the requirements for ship systems and the responsibility of the latter is to produce the desired ships, weapons, and equipment.

The CNO’s needs used to be stated in the form of the so-called "single sheet characteristics”; later on, after review of a preliminary ship design, these were con­firmed in the form of so-called "approved characteristics.”

This process worked fairly well, until combatant ship systems got too sophisticated to allow an adequate performance definition in merely a few "characteristics and the "design-to-cost” concept was imposed by DoD as a mandatory tool to be used in all major acquisition programs.

The lack of a clear definition of operational require^­ments led to frequent misunderstandings which in turn caused change orders, associated upsets in the acquisi^­tion process, and inevitable cost growth.

The return to the Navy of the design of its own ships in 1971 removed the contractual barrier which existed during CF/CD (when a change in requirement meant a change in the contract was necessary) and aga*ⁿ made possible the frequent changes in requirements OpNav is likely to produce during the design period- This necessitated the examination of the whole process of requirements derivation and their translation in^t0 design parameters.

As a result, these problems were identified:

► The "single sheet characteristics” and "approved chat' acteristics” did not define adequately the required ship system performance. The definitions ranged from a m>^x' ture of detailed hardware specifications for radars, weap^­ons, and electronics to the grossest of generalities such as, for example, "best obtainable seakeeping qualities-

► The requirements or performance as required by OpNav and the characteristics in the resulting desigⁿ were unbalanced to the extent that, while certain fun^c­tions of the ships were discussed at great length, other⁵ were only mentioned casually and many were not eveⁿ

discussed.                                                             ^

The two primary reasons for this state of affairs wer- a lack of early disciplined progressive derivation of th^e requirements and characteristics and the lack of a docu^­mented dialogue among the budget managers, the ship operators, and the ship designers.

Figure 3 depicts the chain of events between the tif®^e when the need for a new ship was recognized and the beginning of preliminary design. The multitude ⁰ documents which were required to get the yet und^e' signed ship into the budgetary cycle is indicated at th^e top of the figure. The various organizations whic^ needed to review and approve the program are list^e on the left side of the figure. The major point is concentration of activity above the Naval Materi ^ Command, and the very brief "dips” into the materis side of the Navy for "what if” type information. TW analysis groups in OpNav, while capable of perforrnffif force-level type operational analyses, lacked the techn¹ cal depth in ship engineering needed to assess to ■ sufficient level of detail whether the specified requi^re ments were compatible with the available technolog)' specified equipment, and cost constraints. Therefo^ie'

^ ^en "what if” type questions did come down to the *P designers, they often came with little or no ex­^nation of intent and without sufficient time to re- P°nd ■with adequate technical depth. Distrust devel- _tjj| ^on both sides: on the OpNav side because of _a(j^C ^^arS^e inconsistencies stemming from the lack of 'i^Uate technical depth of the material they received; designers’ side since they didn’t see, nor could ^aPpreciate, the rationale behind some of the "what type questions. It was this back and forth process d_eh . ^u^‘mately resulted in the final characteristics tntion. This definition generally was a mixture be­^en hardware and performance specifications without ’ certainty as to their compatibility. One recent ship suiting from this "method” is the DLGN 36, a ship > as one might expect, many technical problems ^r ^be time of delivery of the ship to the fleet, j ^otb pre- and post-contracting problems can result ,^{0rn (}be deficiencies in the characteristics. Among ^se deficiencies are over-specification in some areas, P cially electronics hardware items, and some omis- ^ns» especially in the definition of required ship sys- performance and the associated environmental _t ^stfaints. An example might be misunderstanding as ^cbe ambient temperature at which the ship’s top P^ecd is expected to be met.

be lack of definitive baseline design requirements

resulted by necessity in decision making by the ship designers. Such decisions were frequently contrary to the real (but unknown) requirements of OpNav. When such contrary decisions were not discovered until after contracting, change orders resulted in increased acquisition costs and production delays. Perhaps worst of all were the cases where designers’ decisions contrary to the real requirements of OpNav were not discovered until after the ships were delivered to the Navy. In such cases either the delivered ships cost too much or their performance was inadequate in certain respects. In some new ships both deficiencies were represented: some subsystems were better and more costly than required while the performance of others was inade­quate.

¹The specific names of these documents, because not germane to the point, ate not spehed out.

ily

°ne of communication. What was required was

Another result of the lack of definitive baseline ^recluirements was the frequent generation of change ^^ers during ship construction by new people in PNav who had ideas about the ships different from , °^se held by their predecessors. Since there was no °^Cunientation of rationale for those past decisions, the new people were unfamiliar with the history of the

^ard decisions made by those who had been there

before.

The basic issue behind these problems was primar­^{an a}uthoritative and definitive document of ship char­^acteristics to satisfy the following needs:

h To define a "contract” between CNO and CNM.

²- To control the design as it progresses to more ^anvanced stages.

T To provide the basis for defining the required Performance tests for the ship.

4.To assure a balanced participation by all those '^vho have a role in the derivation of new ship charac- Tristics such as the ship operators, the budget man- ^a8^ers, and the ship designers.

5.To insure compatibility and traceability between ^rcquirements and hardware specifications.

• REQUIREMENTS:

(a)

• PHASE