F-35 : le rêve aéronautique connaît toujours des déboires
F-35: the aeronautical dream still faces setbacks
7 September 2021
In November 2014, we published a review article on the F-35 and its setbacks (1). Nearly five years later, the programme has made progress, found new buyers, with others indicating that the aircraft had reached its initial operational capability (IOC). But since then, new problems have arisen and raise the question of its real operationality, as its operational evaluation phase is due to end before the end of the year. Here is an update.
The operational evaluation process leading to the long-awaited declaration of the US F-35's IOC should make it possible to validate a number of parameters - an essential point before series production - but doubts about the value of the future announcement quickly arose, before being highlighted by a report from the Project on Government Oversight (POGO). The latter worked in particular on the basis of information transmitted by the Pentagon's Department of Operational Test and Evaluation (DOT&E). In addition to the variously known problems, which clearly took longer to resolve than expected, there were also problems whose scope had been underestimated.
ALIS in the land of maintenance
This is the case for the integrated logistics management system ALIS (Automated Logistics Information System), which must not only allow maintenance - predictive or curative - while establishing schedules, but also the management of spare parts, and this in an automated manner. The approach is that of 'all-in-one' maintenance, whether it is daily or longer term. The development of the system has, logically enough given the scale of the task, been lengthy, but has also suffered from criticisms of its very nature.
On the one hand, access, which is essential for giving the green light to use the aircraft, is provided by a subscription - for a fee - to Lockheed, which retains ownership. On the other hand, the system operates in a network, which requires daily information feedback, also automated, to the manufacturer. This entails the risk of hacking or eavesdropping by a third party, who would then be aware, in real time, of the status of a given fleet. This approach, which was widely criticised, has been partially revised: devices can now be used without being connected to ALIS for 30 days (initially it was 48 hours).
Moreover, the very operation of ALIS is open to criticism, although its use was intended to reduce the burden - particularly the human burden - of maintenance for the client air forces. The process, which is more fluid, should also ensure greater fleet availability. However, in practice, the opposite has been observed. The US Air Force has seen an increase in workload, mainly due to false maintenance alerts, slow network connections, and non-automation, which forces maintainers to manually enter the work done, or even to use parallel databases, which the USAF itself modestly describes as a 'high level of manual effort'. The situation was such that Lockheed itself did not use the system on its lines until 2018 and the US Air Force Secretary could declare on 28 February 2019: "I can guarantee that no Air Force maintainer will ever call his daughter Alice."
Patches have obviously been put in place. In fact, ALIS is said to have gone through 27 versions. But if they can solve some problems, the complexity of the system is such that others are created. Others have sometimes persisted since 2012, without ever having been resolved. This is particularly the case with the sub-programmes. For example, the Squadron Health Management application can declare a device non-operational, while the Customer Maintenance Management System, another application, indicates that it is operational. The future of ALIS and its 65 applications - including some related to planning and debriefing or training (2) - is bleak, so much so that the Air Force has ordered a replacement programme, Mad Hatter, from Lockheed. The idea would be to install the functional parts of ALIS in a cloud. But this will probably not solve everything. For example, the management of spare parts for the various sub-systems of the aircraft, which has never been standardised, even though they come from 1,500 sub-contractors, is causing not only a work overload but also a slowdown in supplies.
A disarmed weapons system?
The F-35 continues to experience several problems with its weapons and sensors. The most serious concerns the EOTS (Electro-Optical Targeting System), which is positioned under the nose and serves as both a laser designator and IRST (Infra-Red Search and Track):
- the first is a maintenance classic: the sensor is permanently installed, thus prohibiting the conduct of air-to-ground missions if it fails, whereas it is sufficient to change pods on more conventional aircraft. In theory, ALIS should allow predictive maintenance to avoid this situation. In practice, however, its poor performance will also affect availability for air-to-ground missions - especially since the purchase of the F-35 came at the cost of reducing the number of aircraft for many air forces;
- The second is even more emblematic: currently, the EOTS does not allow a laser-guided weapon to designate a moving target. The F-35 is therefore not suitable for CAS (Close Air Support) missions. The US Air Force's alternative is to fire GPS/laser-guided ammunition: firing at coordinates must follow a designation either by another aircraft or by a ground team. Tactical flexibility is lost, while the air force may have to acquire the appropriate weapons. Mission scenarios are also made more complex, for example in counter-A2/AD, forcing the deployment of either special forces... or non-stealth aircraft.
The issue of the F-35A's internal gun misalignment has not yet been resolved, especially since it has been found that this problem affects all aircraft in different ways. POGO also states that the tests carried out with the three types of ammunition to be used are fairly unrepresentative and have only involved a few shells of each type in the various employment configurations studied. This criticism is added to that of the calibre and number of shells being smaller than those of the A-10, raising questions about the relevance of using the F-35 as a replacement for the Thunderbolt II, a function that the US Air Force continues to insist on. It also raises questions from the US Army, which is the primary user of the A-10's firepower. Without an EOTS allowing optimal vision shared with ground troops - in particular through ROVER video software - and with a gun that is not very precise and less adapted than that of the Thunderbolt, a real deficit in terms of air support is likely to appear.
There is also the question of the ability to recognise ground-to-air and air-to-air threats. The F-35 must enable its pilots to evolve in an optimal manner according to its dynamic and stealth characteristics, by identifying the threats encountered. To do this, it is necessary to compare the information gathered by the sensors with threat databases called Mission Data Loads (MDLs), without which it is impossible to identify adversary signals - or to discriminate between friendly and allied signals. While these databases need to be updated frequently as the theatre progresses, it took Eglin specialists 12 to 15 months to provide the MDLs used in the F-35's operational assessments, and they have not yet been deemed accurate and relevant. In this case, the software and systems were inadequate and cumbersome. The problem is not just American: Eglin also has to produce MDLs for six F-35 configurations, four Block 4 iterations and 12 geographic regions, amounting to 120 MDLs that will then have to be updated - a task that seems simple at first glance, but could take several months, a timeframe that is incompatible with the deployability requirements of the forces equipped with the aircraft.
The issue of cyber vulnerabilities also remains unresolved, given that the very concept of the F-35 is based on connectivity in several respects. On the one hand, according to POGO, the methodology: the team responsible for testing cyber vulnerabilities is not governmental, but depends on Lockheed, which poses an obvious problem of objectivity. On the other hand, the strength of the F-35 - its connectivity - is also its weakness:
- Within the aircraft itself, the fusion of data from the various sensors (radar, Distributed Aperture System) should provide the pilot with better situational awareness. However, the corruption of one of these systems would contaminate the others, as well as the central computer system;
- The devices communicate with each other;
- the devices are themselves linked to the ALIS system, which is centralised in the United States but connected to the terminals of the user countries. Each potential contaminant entry point is therefore likely to affect the entire fleet;
- It should be added that the F-35s also feed the MDLs automatically, including by transmitting flight profiles and what the sensors "see and hear". Norway became aware of these transfers - which could have been intercepted by neighbouring Russia - which in turn prompted the development of a specific firewall, Sovereign Data Management, at a cost of $26 million.
The platform and the system remain problematic
Even before the start of the operational test phase, the Pentagon acknowledged that the aircraft still had 941 design flaws, 102 of which were "Category 1", i.e. mission-critical and could cause the death of the pilot or serious injury.
Several observers had already pointed out that several Category 1 problems had been redesigned and subsequently considered less serious. The Systems Design and Development (SDD) phase itself was abruptly terminated in April 2018. The tests of the current phase are partly compromised by these problems since the 23 aircraft concerned require an availability of 80%, which seems to be far from being achieved. They are further compromised by the fact that the simulation systems to test a multi-aircraft engagement in a dense air-to-air and ground-to-air threat environment are not yet operational. The simulators themselves do not yet have the flight profiles and relevant information about the F-35 necessary for their operation. This work would presumably be completed by the end of 2019, but these evaluations would then be conducted at the very end of the process.
In addition, structural problems are still observed, again calling into question the test methodology. In this case, the assessment of the aircraft's fire survivability was also entrusted to Lockheed, which determined that it met the contractual specifications in three of the four scenarios considered... whereas the Pentagon's DOT&E estimated the opposite in 2017 (3). More seriously, it was determined that the F-35B and C participating in the tests had structural cracks, so that they would not be able to reach their potential of 8,000 flight hours and might have to leave service in 2026. No information has been provided by the US authorities on the F-35A. Moreover, the marine versions also suffer from very low operational availability: 23% for the F-35B in October 2017, 12.9% in June 2018. In December 2017, the availability of the F-35C was 0%. POGO is similarly critical of the transparency around the communication of the programme, stating that much of the information presented by the Pentagon in previous reports was missing this year. The quality of the evaluation of the aircraft, obviously essential for its operational career to officially begin, could therefore be called into question.
At the same time, however, US timetables are being maintained by the Pentagon's Joint Program Office (JPO), which is simultaneously pushing for the earliest possible withdrawal from service of a large number of F-16s and A-10s. The US Navy's F/A-18C/Ds were retired in February 2019 and the Marines are also being asked to part with theirs. The challenge at this stage is to be able to meet the planned delivery rates - at a time when the US Air Force has been forced to buy the F-15EX, a real wake-up call for Lockheed - but also to meet the promises made in terms of unit cost. The stakes are also commercial: mass production will make it possible to deliver to customers who have been able to set delivery dates in their contracts. However, the problems will not be solved, far from it. For lack of a sound prototyping policy, Lockheed launched pre-production too quickly, adopting a trial-and-error logic that led to the logistical nightmare of successive adaptations "to the right standard" of the first aircraft produced.
However, the stakes are not low: the question of the development of the Block 4/Continuous Capability Development and Delivery (C2D2) standard is already looming. This standard will make it possible to use all the weapons envisaged for the F-35, and in particular the B61-12 gravity nuclear bomb, but its cost remains unknown. Thus, the retrofit of aircraft already delivered is not a given: in April 2019, the JPO estimated it at $10.5 billion for R&D over eight years and $2.5 billion for the retrofit of the 441 American aircraft. For the RAF, the cost will be $486 million, so it is assessing the desirability of the upgrade. In this case, it would qualify its F-35Bs for some weapons (air-to-air missiles, SPEAR 3 missiles). They would also serve as scouts for Typhoon aircraft acting as "ammunition trailers" once they are networked with Lightning IIs. Belgium would receive them directly at Block 4, at a fixed price as stipulated in the contract. In addition to the logistical issues - and all those already mentioned - there are also the more discreet questions about the reliability of the engine.
What is the unit cost?
The budgetary issue remains topical. Although we are now far from the "$50 million fighter" touted in the early 2000s, the price increase was dizzying before starting to decrease in 2018. Belgium's 34 F-35 Block 4s are guaranteed to cost €117.97 million/piece (including two simulators and various services), but the cost for other countries seems higher. Above all, in addition to the uncertainties surrounding the Block 4/C2D2, the stabilisation of costs should also be put into perspective in two respects:
- On the one hand, the programme is not very transparent in budgetary terms. It is difficult to monitor press releases: in addition to batch orders, there are also long lead items for future batches, as well as corrections and anything else that may relate to the motorisation, ALIS or structures. The unit cost thus depends on what is or is not put into the balance: low transparency can thus make it possible to hide a cost increase. Moreover, this issue will be dynamic: for the user air forces, ALIS will thus be the subject of a subscription, so that the unit cost of an aircraft can no longer be "fly away";
- on the other hand, the price is linked to the orders actually placed. There is no shortage of unknowns in Turkey, where deliveries have been suspended, with only six F-35s ordered out of a target of 100; in Singapore, which has ordered four aircraft for testing; in Israel, where some believe the emphasis should be on the F-15IA; in Italy, because of payment difficulties; and even in Japan, after the crash of an aircraft in April, when the fleet's unreliability forced seven emergency landings. It should also be noted that while targets are relatively stable over time, actual orders are still far from being placed...
Conversely, there are still uncertainties regarding new orders (Canada, Netherlands). The increase in costs linked to the Turkish issue does not depend solely on the volume of F-35s finally produced, but also on industrial processes. The United States estimates that 6-7% of the F-35s come from Turkish firms... which will have to be replaced, even if the parts they manufacture from single sources are rare. The fact remains that production costs will no doubt be renegotiated upwards elsewhere, with such discussions leading to additional delays.
In the end, if the F-35 offers the prospect of a digitised aircraft, it must also be noted that this logic is above all that of a managerial dream. Many air forces are expecting a reduction in their maintenance workloads, with unprecedented efficiency gains resulting from information sharing. The prospect of a 'Google view' pointing out the positions of opposing threats and enabling pilots to optimise their flight and stealth is also attractive. But these visions tend to obscure the enemy's side of the equation: both Russia and China are active on the cyber and electronic warfare fronts, while developing sensors to thwart radar stealth, better isolate infrared signatures and locate emissions on which the entire F-35 concept is based.
More broadly speaking, as regards the genetic strategy of the system, the reduction in the number of aircraft, linked to the cost of their systems, is also a vulnerability to endogenous (accidents, maintenance delays) and exogenous (enemy actions, including costly attacks on air bases) attrition. While the problems linked to ALIS, EOTS, data fusion, armament and engine reliability can be settled with dollars, the problem of reducing fleets too much is paid for with less strategic freedom of manoeuvre...