by Peter Yule
adopt Ada. Most adopted close-coupled federated systems, while
Rockwell’s proposal was described as a loose-coupled federated
system. The difference appears to have been that where, for
B U T H O W W I L L Y O U J U D G E T H E M ?
69
instance, Signaal’s close-coupled architecture used a serial data bus
connecting seven multifunction control consoles to two separate
computers, the Rockwell system used a higher-capacity fibre optic
data bus connecting two ‘mass memory units’ to the consoles, each
of which had an embedded processor.27 No proposal achieved the
degree of automated functioning that the navy was seeking.
The Sintra Alcatel proposal was quickly eliminated. This sys-
tem consumed an inordinate amount of power and the lack of any
interface for American weapons was a crucial failing.
Plessey’s Pisces 4, based on the combat system being developed
for the latest UK nuclear submarine, was viewed as risky and of
such complexity that it could not be modified easily. The technical
data in the proposal was thought to be inadequate and the bid was
considered marginal.
The Krupp/Ferranti system was judged to have a significant
degree of development risk but an architecture that made adjust-
ments easier, and was therefore considered to be a fair proposal.
Signaal’s Gipsy combat system was an updated version of
that planned for the Walrus class submarines and used Signaal’s
proprietary programming language rather than Ada. The Dutch
company was considered ‘relatively conservative’28 and did not
hesitate to advise Australian authorities of the specifications it
thought to be overly difficult or impractical. Among these was
that all system functions could be performed at each console
and the same function could be reproduced at each console instan-
taneously. Signaal deduced that this attempt to keep console
databases synchronised in real time would create enormous data
overload as the consoles talked to each other rather than processed
information.29
A poor history of developing Australian industry participation
in earlier projects raised some objections to Signaal, but its pro-
posal was judged to be fair.
Rockwell’s ‘Advanced Combat System’ appealed to the board
because of the familiarity of many of its elements. Singer Libra-
scope’s tactical data management and fire control systems were
developments of the fire control system used in the Oberon update
in 1978. The Thompson CSF sonars were the same as being pro-
posed for the Walrus and Type 2400 and, in any case, the board
felt that if the passive flank array failed to perform, the Micropuffs
sonar from the Oberon update could be substituted. Confidence
70
T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y
in the consortium’s ability to integrate the system rested on what
was an oblique, but unmistakable, reference to Andrew Johnson’s
move from the SWSC to Computer Sciences.30 The Rockwell sys-
tem exceeded the specifications in some areas and also had an
unsolicited feature, the ‘knowledge-based expert system’ for clas-
sifying data.
The board noted that the Rockwell proposal was not a fully
distributed system although it came close to meeting the main
objectives of the concept. It thought that the use of Rockwell’s pro-
prietary expanded service-shipboard data multiplex system (data
bus) was complex and noted that the prototype was still in devel-
opment and would only be produced if the consortium were suc-
cessful. The evaluation recognised the risk inherent in the data
bus but devalued its concerns because of Rockwell’s reputation.
With the benefit of hindsight, the board’s summary findings for
the Signaal and Rockwell proposals are ironic. Although accept-
ing that the arrangement of the Signaal Gipsy system would allow
a ‘graceful degradation’ following malfunctions, it concluded that
‘the use of the close coupled federated system architecture is inher-
ently less reliable than the required distributed architecture’.31 On
the other hand, in its judgment, Rockwell
has designed a new system . . . [that] is a logical follow-on
system to the . . . SWUP combat system. In view of the
technical and management strengths of the consortium
members, the technical risk for the design development and
production of the proposed system is assessed as low.32
The board judged Rockwell’s proposal to be strong to very strong.
Before completing its assessment of the combat systems, the
board looked at the systems on the Walrus and the Upholder and
judged both to be inadequate. The purpose of this exercise was
to test the navy’s position that existing combat systems could not
perform better than the upgraded Oberon system. Confirmation of
this position reinforced the navy’s conviction that the intellectual
base for its combat system concept was sound and contributed to
a misplaced presumption that the success of the Oberon upgrade
program could easily be replicated.
The technical evaluation of the proposals for submarines
and combat systems resulted in clear winners: IKL/HDW’s Type
2000 submarine and Rockwell’s ‘Advanced Combat System’. Yet
B U T H O W W I L L Y O U J U D G E T H E M ?
71
technical merit was only part of the assessment – the evaluation
teams also looked at cost, support and Australian construction.
The French, German and Swedish submarine designs were sig-
nificantly cheaper than the Dutch and British. The French put in
the lowest bid but the lack of supporting documentation raised
doubts about its veracity. The second lowest bidder was Thyssen,
with IKL/HDW and Kockums’ bids lodged at prices only two and
five per cent greater, respectively, than the French. In contrast,
the Walrus and the Type 2400A were 33 and 46 per cent more
expensive respectively.33
The price differences mostly arose from inefficient Dutch and
British building practices. The labour hours needed to build the
British submarine were six times those for the German subma-
rine. Even on an hourly rate of A$14.90, British workers could
not compete with their German counterparts on A$25 an hour.
The Thatcherite revolution that transformed the British economy
during the 1980s had not yet reached the Vickers shipyards.
Price was not a major issue between the combat systems as the
acceptable systems varied by no more than 13 per cent, at prices
around $500 million.
The Defence Department had allocated $1.546 billion for the
project, yet nothing in the matrix of options met that price: the
favoured combination of the IKL/HDW submarine and the Rock-
well combat system came in at a little under $2 billion. Buying
an option already under development was no more affordable, as
the Upholder and the Walrus were the most expensive proposals.
After some debate it was decided that the IKL/HDW Type 2000
to operate than the Walrus or the Type 2400A Upholder. Only
the IKL/HDW bid fully met the requirements for logistics sup-
port, although Kockums’ proposals were considered attractive.
Kockums maintained the Swedish Navy’s submarines and, con-
sequently, had considerable data on submarine reliability with
which to reduce support and maintenance costs.
Most of those who saw the project as a catalyst for indus-
trial regeneration believed that only the Germans and Swedes
really understood the project’s industrial requirements, and many
of them favoured Kockums for its modular construction tech-
niques. The evaluation board agreed that Kockums’ approach
would encourage involvement by a wider spectrum of Australian
72
T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y
industry than would its competitors.34 Kockums’ production
methods were judged to be the most modern of the contenders
and it was the only company that used computer-assisted design
and production management technology.
The Dutch bidders, RDM, did not appear to realise that both
the submarine project team and the Australian Labor govern-
ment were fully committed to achieving genuine Australian indus-
try involvement in the project, and its production planning only
allowed for an Australian content of 60 per cent by the end of the
program. They had teamed with Lend Lease, a conglomerate that
controlled Civil & Civic, then Australia’s largest property devel-
oper. This might have given the Dutch a capacity to develop con-
struction facilities, but the resulting consortium would have been
entirely reliant on RDM and its traditional construction methods
in building the submarines.
Similarly, Vickers showed little enthusiasm for involvement
with Australian industry. The tender request looked for proposals
that would cover building the submarines, integrating their sys-
tems, developing industry facilities and providing long-term sup-
port, but Vickers tendered only for submarine design and building
the first vessel in the United Kingdom. It left the Common-
wealth to negotiate separate arrangements for any construction
to be done in Australia. The board was not impressed, fearing
a divided responsibility between designer and builder that gave
the ‘potential for disruption, delay and cost increases of the type
experienced with the patrol craft and [other] projects’.35 It also
‘noted with concern the reluctance of the tenderer to enter into
the preferred contractual arrangement’.36 In any case, the board
assessed that ‘production technology currently used in the UK
is . . . conservative and labour-intensive’.37
The British compounded the perception that they were
opposed to building in Australia by emphasising in their inten-
sive advertising and lobbying efforts the importance of maintain-
ing the Australian navy’s traditional ties with Britain and making
it clear that they were doubtful about the feasibility of building
submarines in Australia.38
By early 1985 the Tender Evaluation Board was ready to
pick the winners. The technically unsatisfactory responses were
excluded. The marginal designs were examined for non-technical
benefit. Being only four per cent more expensive than the
B U T H O W W I L L Y O U J U D G E T H E M ?
73
Type 2000, the Kockums bid warranted further consideration.
Having no price advantage, the Vickers bid was rejected.
In separating the Swedish and Dutch bids the board noted that
the cost to rectify deficiencies was much less for the Type 471
than for the Walrus. The board had become accustomed to the
novelty of a Swedish design and changed its attitude. The Wal-
rus ‘offered compliance with a reasonable proportion of the RAN
requirement but had a number of notable deficiencies’.39 In con-
trast, the Swedes ‘offered a high degree of capability relative to
the RAN requirement by providing the required endurance, speed,
depth, manoeuvrability and had only minor shortfalls in most
characteristics’.40
The Walrus remained less risky because it was the closest to
the original requirement of being in service with the parent navy,
but this was offset by Kockums’ superior proposals for Australian
industry involvement. The board had ‘great difficulty in reaching
a common view on the value of these different attributes’, but
ultimately it preferred the superior efficiency of Kockums’ building
processes and the consequent price advantage. The board also
believed that the Kockums proposal ‘would be more conducive to
effective competition with IKL’.41
When the board came to decide between the combat system
proposals, Plessey was quickly excluded. Even though it was the
cheapest acceptable bid, there were doubts about its technical risk
and the basis of its costings. Choosing between the Signaal and
Krupp proposals was more difficult. Both provided similar levels
of performance but the Krupp system was marginally cheaper.
However, the price margin was slight and Dutch navy support
for the predecessor system in the Walrus reduced the technical
risk of the Signaal system. Further, the Dutch were familiar with
American weapons and this tipped the balance in their favour to
compete with Rockwell in the next stage.
Nonetheless, the board had doubts about the competitive
project definition study, for it would demand many resources
that Oscar Hughes and Harry Dalrymple thought might not be
available.42 The board also feared that the time required for the
project definition study would jeopardise the Oberon replacement
schedule.43
Essentially, the IKL/HDW submarine and the Rockwell combat
system were thought best by a ‘substantial margin’.44 However,
74
T H E C O L L I N S C L A S S S U B M A R I N E S T O R Y
recognising the government would consider further development
with only a single submarine designer and systems developer too
risky, a definition study involving the two submarine designers
was proposed as a ‘reasonable compromise between the Com-
monwealth’s interests and its limited resources’.45 This would sub-
ject two-thirds of the project costs to competition. Ultimately, the
board felt that the Rockwell combat system proposal was not only
clearly superior, but it was the only one that was ‘a logical and
evolutionary development of the Oberon combat system’, making
a competitive project definition study unnecessary.46
There was a final twist in the tale in early 1985 before the
winners were announced, as a rearguard struggle took place in
Canberra to push the project off its path. The civilian policy mak-
ers led by Alan Wrigley still argued that the navy had not justified
its proposals, and as late as May 1985 the defence force develop-
ment committee – the last stage at which departmental ad
vice was
considered before being sent to the Minister – recommended that
a price ceiling be placed on the project. The recommendation was
not accepted.
Jim Duncan and other missionaries for the project believed
that the bureaucracy was ‘deliberately delaying consideration of
the project and threatening to take the process beyond the tender
validity date’. This would have meant recalling tenders, thereby
‘giving the increasingly strident Vickers lobby another shot at get-
ting their submarine selected’. Duncan recalls that several major
projects had been delayed by clashes between civilians and uni-
formed personnel on the defence committees and the submarine
project could similarly have been swallowed in the morass of
departmental in-fighting. He saw Kim Beazley’s direct links with
members of the project team like Graham White and Rod Fayle
as being crucial in breaking the deadlock and forcing decisions to
be made.
Eventually, on 9 May 1985, the submarine project went to cabi-
net. The defence civilians still argued against the risks of technical
innovation and Australian construction, while the project team
presented the strategic and operational case for large submarines
and the logistic and industrial arguments for building in Australia.
Crucially, they were supported by the Chief of Naval Staff, Admi-
ral Mike Hudson, and cabinet agreed with his conclusions. The
civilians had lost the argument; there would be no reconsideration
B U T H O W W I L L Y O U J U D G E T H E M ?
75
of options and the project would be for six submarines, with an
option for two more. All would be built in Australia. The only
area where cabinet did not agree with the navy’s proposals was
on the evaluation of the combat system: cabinet thought it pru-
dent to have a backup option and ordered the inclusion of two
combat system suppliers in the definition study.
There was no surprise when Rockwell and Signaal were chosen
for the combat system contest, nor with the choice of IKL/HDW
for the submarine contest. The selection of Kockums as the other
submarine contestant was a shock.
C H A P T E R 8
Spies, leaks and sackings: from tender
evaluation to project definition study
While the tenders for the submarines came from individual Euro-
