In recent years, the cost of warships has soared, making a hull of about 4000 tonnes the largest many navies can afford to buy in quantity, in terms of initial monetary outlay, dockyard construction facilities and maintenance. The frigate (and destroyer for that matter) are medium sized ships, that tend to fulfill a 'jack-of-all-trades' role within a fleet, having to counter air, surface, missile and submarine threats.
The development of the successor to the Leander class frigates proved to be a long drawn out affair, and eventually became the Type 82 fleet escort destroyer which was designed to act as an escort for the CVA-01 fast carrier. With the cancellation of this project and with only one out of the five Type 82 vessels (HMS Bristol) being completed, and the decision to phase out the remaining aircraft carriers, by the mid-1970s, the Royal Navy would have no effective air defence vessels. Top priority was given to the design of an air defence escort vessel, and the result was the Type 42 destroyer, the first of which was HMS Sheffield. With the priority given to the Type 42, the Type 21 was brought in as a stopgap measure, and work on the Type 22 proceeded slowly. The first of class was eventually laid down in February 1975, with the others proceeding at yearly intervals. The specification called for the ability to conduct sustained anti-submarine warfare, but have a general capability as well. An early decision was made to have the same engines as the Type 42 destroyers, which meant that the Royal Navy would benefit from economies of scale, and operate a common spares holding.
The main anti-submarine weapon was the Lynx helicopter (of which there were two with a double hanger) and together with the Stingray torpedo, formed a capable system. The design of the hull closely followed that of the Leander class, which was renowned for its sea-keeping ability. The hull is flush-decked with the forecastle being raised, as in the Leander ships. A superstructure deck stretches nearly the full length of the ship and provides a generous amount of internal space for its electronic systems. The upper superstructure was kept as clean as possible with only the necessary systems visible, to help with cleaning down nuclear fallout, and to minimise the surfaces a radar signal could lock onto. The ship was armed with four container Exocet launchers, two sextuple Seawolf missile launchers, two 40mm Bofors guns and the Knebworth / Corvus multiple rocket launcher. There is an extensive range of electronic and radar systems, including the Type 967/968 radar, providing information for the Seawolf tracking radar and Exocet guidance system, and the Type 910 tracking and guidance radar systems, a comprehensive ECM outfit, and a Type 1006 navigation and short range surveillance radar.
The first of class, HMS Broadsword, was laid down in February 1975, and commissioned May 1979. From the fifth of class onwards, it was decided that the hull would be lengthened by 41 feet to provide more internal space, for munitions and additional sonar equipment. Boxer was the first of the Batch II ships, and as well as the additional space, had a Computer Assisted Command System (CACS-1) fitted, which was intended to replace the Computer Assisted Action Information System (CAAIS), but needed a great deal of work just to get it up and running.
Although the original intention was to build a large number of Type 22s, the 1981 Defence White Paper cut the programme back as they were deemed too expensive, and only one more was to be added to the six already approved by the time of the White Paper. However, the Falklands conflict saw Brilliant and Broadsword assigned as close escorts to the carriers Hermes and Invincible. The ships emerged from the campaign with excellent reputations, and the Seawolf missile system had proven a great success. The Seawolf was completely automated, with no human operators, and on many occasions "the first warning of impending attack was the 'whoosh' as a Seawolf left its launcher." [note 1]
As a result of the conflict, it was announced that a further five Type 22s would be ordered to replace the losses suffered in the campaign. Originally, three of the five ships would be of a new Batch III version. However, as a result of delays to the Type 23 programme, a further two Batch III ships were ordered and earlier details revised, so that overall, four Batch I, six Batch II and four Batch III ships were finally completed. Alterations to the design of the later ships, included the Mk 8 4.5in automatic gun, an enlarged hangar to house an EH101 Merlin or Sea King helicopter, and a new Type 911 tracking radar for use with the Seawolf GWS25 system. The previous Type 910 radar systems had been found to have difficulty tracking targets very close to the sea due to 'second path' reflections. The new radar used higher 'K' band frequencies, thereby reducing the problem. HMS Brave was also used as a test bed for two Rolls Royce Spey SM1 gas turbine engines which needed to be tested in an operational environment before being fitted to the Type 23 frigate. While the remaining Batch II ships kept the Olympus / Tyne configuration, the Batch III ships had the Spey / Tyne arrangement in which both engines could be coupled to the shafts simultaneously. Additionally, the weapons suite of the Batch III ships was enhanced with the 4.5in gun, replacement of the Exocet missiles with the Harpoon, fitting of the 'Goalkeeper' Close In Weapon System, which is designed to engage incoming missiles from 1500 meters down to 350. They also carry the STWS-2 A/S torpedo system, two LSE 30mm guns and have been fitted with two GSA-8 Sea Archer 30 series Electro-Optical directors.
The late Batch II and Batch IIIs were fitted with CACS-5, as fitted to HMS Boxer and destined for the Type 23s. Despite the problems, the Type 22s remain some of the most heavily armed ships in the Royal Navy, and demonstrate the increase in the size, capability and cost of modern frigates. The ships ordered in the wake of the Falklands conflict were finally completed by 1989, but there was a gap of several years before reasonable numbers of Type 23s came on line. The Royal Navy should count itself lucky to have been able to secure "so many of these fine ships although it was unfortunate that it took a war in the South Atlantic to provide the incentive to build them." [note 2]
The Type 23 was originally envisaged to act as a platform for a towed sonar array, carry light missile system against air attack and have facilities to land and refuel helicopters, although it would not carry a hangar. This was basically to keep the unit cost to around £70 million and in many ways, resembled the philosophy behind the limited capability Type 14 frigates of the 1950s. In order to achieve the lowest acoustic signature to allow the towed array sonar to function successfully, there was detailed research into suitable hull forms, and a novel propulsion system - Combined Diesel Electric And Gas Turbine (CODLAG). The system had three separate elements - gas turbines for cruising and high speed, and diesel generating sets driving electric motors for low speeds and quiet running. In the end, the Rolls Royce SM1A Spey gas turbine was chosen, and although it only produced 18,775 shp as opposed to the 28,000 shp in the TM3 version of the Olympus engine, great efforts were being put into keeping the ship to a length of 100 meters and displacement to around 2,500 tonnes. The armament initially consisted of one OTO Melera 76mm gun, Exocet MM40 missiles and two STWS-2 triple AS tubes. Although this represented the 'minimum frigate', and may have been built for something fairly near the target price, it would have been limited in its operational applications.
Even before the Falklands campaign the design was recast, lengthening the hull by 15 meters, adding a hangar for helicopter operations, light automatic guns and the Seawolf missile system. The ship now began to approach the Type 22 for cost and complexity (£90 million as opposed to £120 million), and these were not yet the last changes that would be made. As a result of the Falklands conflict, modifications were made to damage control arrangements, and the ship was divided into five self-contained fire control zones, each with their own fire-fighting equipment, escape routes and electrical power supply. New fireproof materials and non-toxic substances were incorporated into the design, and many areas were armoured against shrapnel damage. The Vickers 4.5in gun replaced the OTO 76mm, and a vertical launch system for the Seawolf missiles was included (GWS26), instead of the GWS25 six-round launcher previously. Two additional fixed torpedo tubes were added, two single 30mm guns were installed on mountings abreast of the funnel and the bow-mounted Type 2050 sonar replaced the hull-mounted Type 2016. Length thus increased to 133 meters overall, and displacement (fully loaded) rose to 3,100 tonnes.
Some of these decisions caused a series of political battles, including selection of the surface-to-surface missile system (the McDonnell-Douglas Harpoon was finally chosen over the Italian Otomat, French Exocet and British Aerospace Sea Eagle) and the choice of the tracking radar for the Seawolf missile system (originally intended to be the Hollandse Signaalapparaten VM40 tracking radar, but the GEC-Marconi 805SW I-band radar won the tender). To this was also added the eventual cancellation of the CACS-4 and the tendering for a new command system, which meant that the early Type 23s were without a computer command system at all. A British consortium (which included Ferranti) won, and the new system was based on Ferranti's FM2400 computers which were designed to replace the older FM1600s in the Type 21 and 23 frigates. On top of this, the Government was slow to place orders despite the importance of the Type 23 to the Royal Navy. This caused apprehension to among those dockyards that were looking for work and those in political circles seeking a commitment to maintain a 50-ship destroyer and frigate fleet. Despite the various technical and political problems that have surrounded the Type 23 frigate, the final ship is a well balanced design but also an excellent anti-submarine platform.
Following on from this, an official NATO Staff Requirement was drawn up and presented to each member country by mid-1987, after which there was a period of consideration before a Memorandum of Understanding (MoU) could be signed to then proceed to the Project Definition Stage. Although it seemed likely at this stage that the British Government would withdraw from the project (due to cost and incompatibility with Royal Navy Staff Requirements), a major argument against such an action was that British companies would then almost certainly be excluded from bidding for contracts to supply equipment. Therefore it was announced in early 1988, the UK would participate (at a cost of £100 million) in the Project Definition Stage and Admiral Geoffrey Marsh RN would be Project Manager. Unfortunately, late in 1989, the UK declared that it would no longer be continuing in the NFR90 project, and that it would pursue a national replacement for the Type 42 destroyers at the end of the 1990s.
The NFR90 project offered a great opportunity for European defence collaboration and would have been a significant symbol of allied unity. Unfortunately, there was difficulty in reaching agreement on work sharing, which often conflicted with the goal of cost-effectiveness. Also, approval from each participating nation was for one stage only, and all work would have to stop while the following stage is negotiated. A British, nationally designed frigate could take up to eight to ten years; the NFR project could have lasted fifteen or twenty. There was also the problem of designing a ship to accept weapons that do not as yet exist, especially when the participants do not want the same weapons fit. While there was a Project Office in which the representatives from each country were responsible to the Director, as well as seeing that their particular national requirements were met. They also had to organise a consortium of many different firms from the participants, many of whom were not keen on sharing trade secrets.
Even at that early stage, it was recognised that the allocation of work sharing would be a delicate matter. The problem was seen as the fact that all three partners were capable of building a nationally designed frigate themselves and with no-one being a specialist in one area, or lagging behind the others in another, there was no natural division of labour. It was therefore envisaged that the work would be done in 'packs' by the naval yards and then proceed according to a 'building block' approach. Additionally, a major obstacle to the project proceeding on schedule was the disagreement between the UK and its continental partners over the radar configuration for the frigate's Principle Anti-Air Missile System (PAAMS). The UK wanted a MESAR variant, while the Italians and the French being happy with the EMPAR system. As a compromise, the three governments asked two consortia, Eurosam (Thomson-CSF/Aerospatiale/Alenia) and UKAMS (BAe Dynamics - GEC Marconi) to conduct studies to see if a single interface could be produced that could handle either type of radar. By the end of October 1994, the question of whether to adopt the EMPAR or MESAR/SAMPSON variant had been effectively left up to the UK Government as the defence industry had guaranteed that both EMPAR or SAMPSON could be integrated in to the PAAMS architecture. One defence industry source described PAAMS as the main hold up in the project - with the UK's insistence on having a MESAR/SAMPSON radar system, this could well lead to a break-up in the programme. Even, the IJVC's managing director commented on the unusually high level of risk, especially given the constraints within which it had to work - the tender would only be accepted once the three governments had signed the Supplement to the July MoU. This would release the money for the definition phase, but up until that point, all the costs would have to be met from the companies' own resources.
Finally, in February 1995, the IJVC was formally established in London, the national prime contractors being DCN International, GEC-Marconi Naval Systems and Orrizonte SpA. The new consortium was supposed to have been formed in the previous October, but a dispute over commercial conditions with the three National Armament Directors delayed its establishment. This strained relations with the Government-level Joint Project Office, which was unable to begin actual procurement without its industry counterpart. The situation was resolved when the IJVC Horizon partners received assurances that non-recurring costs incurred in setting up the joint venture would be considered as a through-life overhead, recoverable at a later stage. By this time, concerns had been voiced over the possible gap between the CNGF requirement and the level of funding available. According to one source, IJVC Horizon would have to spend double what it had originally intended to produce all the deliverables currently agreed.
By the middle of 1995, CNGF was the largest surface warship project in Europe, and exerted influence on procurement policies all over Western Europe. But the complexities of European defence collaboration were starting to have an impact and a report from the National Audit Office thought that the MoD had underestimated the overall procurement cycle by up to four years. It also warned that care would be needed if work share and cost problems were going to be avoided. While being reasonably settled in terms of management and leadership of the industrial partners, the three Governments seemed to have difficulty in agreeing certain aspects of the vessel and the main weapon system. The delays in signing the MoU for the PAAMS meant that the remainder of the programme was delayed, including the start of the definition work on the combat management system (CMS), electronic warfare system (EWS), the integrated communications system (ICS) and the warship platform. The UK effectively put a hold on further progress by refusing to sign the Supplement 1 to the Project Horizon MoU until the three partners had reached agreement over PAAMS in that they should find a cost-effective technical solution that also met national work share arrangements. The UK also had a desire to derive maximum long-term benefit from the radar for PAAMS and had reservations about the performance of EMPAR. Issues of work share, cost and competition complicated the final decisions over two major PAAMS subsystems - the Long-Range Radar (LRR) and vertical launch system (VLS).
Three MoUs were finally signed in March 1996. They covered the general rules governing the three partners' collaborative effort for overall development and production of the PAAMS programme (PAAMS MoU), the PAAMS Full Scale Engineering Development Initial Production Phase (PAAMS MoU Supplement 1) and a supplement to the Horizon programme covering the design definition phase (CNGF Programme MoU Supplement 1). Despite this, many commentators commented on the continuing delays and arguments, which would slow the programme down, over differing national requirements, the choice of electronic systems and defensive weapons. As 1996 progressed, the PAAMS related delays and the hold up in the award of the SAMPSON development contract meant that the probable in-service date again slipped to sometime in 2006. This meant that there would be growing problems in maintaining the Royal Navy's Type 42 destroyers, as well as the French and Italian Suffren-, Cassard-, Doria- and Audace-class anti-air warfare ships, as the earliest Type 42 destroyers would be approaching thirty years old, HMS Birmingham being commissioned in December 1976. The cost of continuing to maintain these ships was regarded as considerable, as was the cost of the Sea Dart missile system.
As the programme continued, things were reluctant to improve. There were disagreements in early 1997 over the type of vertical launch system to be employed. The French and Italians had favoured the Franco-British-Italian Sylver A50 developed by DCN, Alenia and BAe Dynamics. The UK however, had shifted to support the Mk 41, supplied by Lockheed-Martin, which would allow the installation of US SM-3 based theatre missile defence missiles. The UK had also been refusing to negotiate a full-scale engineering development and initial production contract for PAAMS on the terms that had been offered by industry to the PPO, and had resisted moves by France and Italy to relax the PAAMS performance specification set out in the original agreement. This continued uncertainty had forced the JPO to shelve a number of key design definition contracts. What was worse, the divergence of performance goals coupled with different national contracting procedures had produced a schism in the programme, which had been made public by a leaked letter from the Chief of Defence Procurement (CDP), Sir Robert Warmsley. The letter was addressed to his French and Italian National Armament Director (NAD) counterparts, and outlined the UK's concerns - that the UK still wanted Column 2 performance, which was the Royal Navy's minimum acceptable performance criteria, meaning a local area capability against simultaneous threats. France and Italy had been ready to accept Column 1 performance, which was really only a replacement for the existing SM-1MR missile. Warmsley had also noted that industry had been unwilling to respond to the joint requests either in the form of an acceptable offer for PAAMS or more information.
Finally, in mid-1997, the NADs of France, Italy and the UK conditionally endorsed the industrial management framework for the collaborative development and production of PAAMS. PAAMS was designed to provide CNGF with area, local and point defence capability, based on Astor 15 and 30 surface-to-air missiles. The Astor 15 missile had just been successfully tested in a 'hit-to-kill' engagement with a live MM38 Exocet anti-ship missile, demonstrating that it could carry out the local-area defence mission, against a crossing target, as required by the Royal Navy. Later that year, two consortia were awarded Phase B Project Definition study contracts to develop proposals for the fully integrated EWS. The two groups involved were the ARCEO team (Racal-Thorn Defence with Alenia Difesa and CS Defense) and the JANEWS consortium (Elettronica, Horizon GIE - a joint venture between Thomson-CSF and Dassault Electronique - as well as GEC-Marconi). They were awarded £4.5 million each to conduct an eighteen-month programme of definition work. The EWS was designed to be a fully integrated and autonomous subsystem, with electronic support measures (radar-band), onboard jammers and offboard decoys, coordinated by advanced software. The specifications of the system were ambitious, with high sensitivity, high bearing accuracy, modulation on-pulse analysis and programmable high-radiated-power jamming among those deemed necessary. The initial Phase A Project Definition work had been done in 1994 and identified point defence jamming as a potential problem area. Italy favoured the cross-eye technique (using solid-state transmitters) while France and the UK looked towards the use of offboard decoys for angular seduction. It was also decided that PAAMS would be procured via a 'twin-track' approach due to the differing contracting requirements of the UK and France/Italy.
By the summer of 1998, the industrial partners developing PAAMS were at the point of being ready to sign an agreement on full-scale engineering development and initial production. The PAAMS Programme Office had virtually completed its deliberations concerning the selection of outstanding elements of the PAAMS system, as well as the allocation of work shares, while the UK Government had nearly finished its Strategic Defence Review. Aerospatiale had already begun extolling the virtues of the PAAMS system as a unique multi-mission air-defence system, which uses a single active-homing missile and sensor system to fulfill the self-defence, local-area defence (over seven kilometers) and area defence (up to fifty-five nautical miles radius). Combined with a radar that has a sufficiently high update rate (such as SAMPSON), Aster could be used in the theatre anti-ballistic missile role, but it would need both a new warhead (to produce larger fragments) and greater range. Royal Ordnance had also unveiled proposals for its medium-calibre gun system and hence went head-to-head with Italy's OTOBreda. The requirement called for a 5 in/127mm weapon, with gun mount, control system, magazine, hoist structure and handling system, a computer system and munitions. OTOBreda had offered its new 127mm/54 Lightweight Gun Mount (in partnership with VSEL Armaments) while Royal Ordnance's offered the Mk 45 CNGF Gun Weapon System (GWS), which was almost identical to the US Navy's new Mk 34 Aegis GWS. It was the only system at that time to be purposely designed to take advantage of the new EX171 extended-range munition being developed by Raytheon TI Systems. Finally, the PAAMS Programme Office chose the Sylver missile launcher (made by DCN) rather then the rival Mk 41 vertical launcher, which is manufactured by Lockheed Martin. This was something as a blow to the Royal Navy as they had expressed interest in acquiring a theatre ballistic missile defence capability based on the Standard Missile Block IV and the Tomahawk Land Attack Missile. The Sylver launcher, as it was originally configured, could not fire either of these missiles. With the impasse over PAAMS seemingly cleared, attention turned to the Project Horizon programme, which covered the ship and all other combat systems.
A second grant was made for further design definition work to the prime contractor, Horizon IJVC, for outline design work on the ship, combat system and subsystems and the selection of marine engineering. Supplement 2 to the Project Horizon MoU was due to be signed late in 1998 and would have paved the way for a detailed design and FOC build contract award to Horizon IJVC sometime in 1999. Adhering to this schedule was seen as vital to stay in touch with the 2004 in-service date. However, budgetary problems had already started to loom, with the partner countries starting to demand a reduction in the size and cost of the CNGF before committing to it. Work nonetheless continued on assessing the rival bids for the Combat Management System and the Fully Integrated Communications System. EUROCOMBAT (BAeSEMA, Thompson-CSF and Alenia) and HEPICS EEIG (Marconi, Matra Defense, Dassault Electronique, Matra-CAP Systemes, and Datamat) were contesting the CMS competition. The FICS competition was being fought by NICCO Communications (a limited company comprising Thompson-CSF, Redifon MEL, Marconi Communications and Elmer) and a rival team comprising Italtel, BAe Defence Systems, Racal Communications, Bull SA Europe and Dassault Electronique. Bids for the Electronic Warfare System had been in by the previous September and the two rival EWS consortia are ARCEO (comprising Racal, Alenia and Matra Defense) and JANEWS (comprising Electronica, Thompson-CSF, Dassault Electronique and Marconi Electronic Systems).
Finally, in April 1999 the UK decided to pull out of the platform component of the Common New Generation Frigate, although it confirmed it would still continue with the PAAMS component, despite MoD assurances that the decision had been taken by all three partners. "In suggesting that its termination was a joint decision, we judge that the MoD is putting a rather favourable gloss on the programme's failure" and that while such a decision was formally put forward as a joint one, from "comments we have received from France and Italy, it is quite clear that this was the initiative primarily of the UK, which withdrew from the project after losing patience with the programme's unfocused management and the high price of the warship." [note 3] This effectively marked the end of Project Horizon, with the French and Italians looking at the possibility of bilateral cooperation, and the UK having decided on a national warship programme. The decision to pull out was eventually arrived at due to the UK's continuing dissatisfaction with the industrial and management structure of the collaborative project, despite trying to stimulate the process by refusing to sign the PAAMS agreement. The UK wanted the prime contractor designate, IJVC Horizon, which was owned in equal shares by DCN International, Orrizonte SpA, and GEC's Marconi Marine, to reconstitute itself with Marconi Marine taking on the single prime contractor role and full responsibility for price, performance and delivery. Further difficulties arose when France insisted that DCN should take responsibility for the combat system within this new structure. Even "the firms involved in the Horizon IJVC could not agree amongst themselves on how such a prime contractor should be established, nor which one of them should become the leader." [note 4] There was additional disagreement over the combat management system (CMS) with France insisting that the system should be based around a HEPICS CMS, drawing on core technology from DCN's SENIT 8 system. The UK preferred the rival Eurocombat CMS on both price and performance. A problem also arose in that there seemed to be a lack of separation between the 'customer' and the 'supplier' as regards the CMS as one of the consortia bidding for the contract (HEPICS) contained DCN of France which was also in the Horizon IJVC. Thus "this blurred the contractual boundaries considerably, with DCN putting itself in the awkward position of being part of the team of one of the bidders while it would also be involved in awarding the CMS contract." [note 5]
The Phase One design definition studies for Project Horizon will be used by all three nations to determine their national procurement strategies.
It was eventually agreed that the participants could employ the radar of their choice, even though this meant a departure from the principle of utilising common equipment throughout the design. Not only that, but the different navies effectively operate in different environments. The Italian Navy is essentially concerned with the shallower, more confined waters of the Mediterranean and Adriatic Seas, while the French Navy looks to operate around the Bay of Biscay, the littoral Atlantic waters near to Europe and the Mediterranean as well. The Royal Navy however, needs ships that can operate in true deep-water conditions, at mid-Atlantic range. So, while France and Italy would have been happy with a ship of around 3 - 4,000 tons, the UK could not really make do with anything less than 6,000 tons. "A compromise was reached in definition of a common operational requirement but it was only a compromise - with all that word implies." [note 7] While the joint project office (JPO) that had been formed in London was representative of the Ministry of Defence of all three nations and the commercial shipbuilders that had come together for the project, it was only really a cipher. The JPO had no real authority, in that it required consent from all three governments to make decisions of any importance. In the end, national interests overrode common goals. There has also been strong criticism of the slow decision-making process and that because no single production line was envisaged, it would have been difficult to realise a reduction in hull production costs and collaborative overheads would probably increase programme costs. This negated the entire reason for a collaborative project. There was additional concern being voiced from British industrialists in that although the UK was paying one-third of the development costs, it would actually end up receiving a low share of the work involved. While Rolls Royce was very likely to end up the winner of the contract to supply propulsion plants, because of juste retour considerations, there would be little left over to go to the electronics industry. The French and Italians were seen as suppliers of sixty-six percent of the equipment on each of the British ships and since the equipment was to be the basis for other UK warships, British industry envisaged itself being excluded from other UK programmes while French and Italian competitors won contracts at their expense. There were also problems with regard to the industrial structure involved with a consortium of firms from each country. The "unresolved turbulence in the European defence manufacturing industry" and the "hodge-podge of interlocking companies and project teams for the programme, with different reporting lines, also illustrates the complexities of trying to bring off a collaborative project. It does not, we judge, provide a very positive pattern to be followed in the future." [note 8] GEC-Marconi was particularly unhappy with the management arrangements in that "the structure of the Horizon International Joint Venture Company (IJVC) was sub-optimal. Shareholders in the IJVC taking shared liability on a high-risk programme inevitably required the protection of agreement by unanimity. This led to a lack of clear leadership and the shareholders tended to be more concerned with protecting their domestic interest particularly in terms of work share rather than managing the project on an objective basis." [note 9]
Horizon's death knell should sound a warning to those who would sacrifice national defence requirements in the interest of international collaboration. That is not to say that collaboration cannot be advantageous, for example in allowing some nations to implement programmes that they could not afford alone, via the spreading of costs and encourages the sharing of technology, cross-border industrial alliances, joint ventures and partnerships and adds closeness to partner nations. Collaboration is an ideal way forward in an era of ever tightening defence budgets, but at times, idealism must make way for realism. Also, some benefits have emerged form the Horizon project, such as the PAAMS programme, and the electronic warfare and integrated communications systems that could still be applied to other warship programmes. The lessons that must be learnt for future collaborative projects are that: