The Director of Operational Test and Evaluation (DOT&E) recently released a scathing assessment of the F-35 program as part of his annual report. Buried inside 48 pages of highly technical language is a gripping story of mismanagement, delayed tests, serious safety issues, a software nightmare, and maintenance problems crippling half the fleet at any given time.
The report makes clear just how far the F-35 program still has to go in the development process. Some of the technical challenges facing the program will take years to correct, and as a result, the F-35’s operationally demonstrated suitability for combat will not be known until 2022 at the earliest. While rumors that the program office would ask for a block buy of nearly 500 aircraft in the FY 2017 budget proposal did not pan out, officials have indicated they may make such a request next year. The DOT&E report clearly shows any such block commitments before 2022 are premature.
The report’s candor about the airplane’s problems is unique among the DoD’s other reports about the performance of the F-35. It only exists because Congress created an independent operational testing office in 1983 to report only to the Secretary of Defense and Congress. Without this office, significant F-35 problems might never be revealed until failure in actual combat.
As damning as this report is, the Joint Strike Fighter (JSF) Program Office quickly issued a statement disagreeing with the report’s emphasis—but acknowledging that every word of it is “factually accurate.”
The F-35 program is already years behind schedule: the first plan was to have the initial batch of the aircraft available for combat in 2010 and deployed in 2012. This report shows timelines slipping even more.
Crucial weapons delivery accuracy tests (WDA) serve as a good example. The weapons test events are important because rather than just testing to make sure an individual component functions properly, they test the entire kill chain, “the complete find-fix-identification (ID)-track-target-engage-assess-kill chain for air-to-air and air-to-ground mission success.” This means the tests will see if a pilot can locate and properly identify a target, hit it with the right weapon, and then tell if the target has been destroyed—just the sort of thing a pilot would have to do to be effective in combat.
The Joint Strike Fighter Operational Test Team (JOTT) identified 15 WDA tests for the Block 2B aircraft that the Marine Corps declared ready for combat last year. Twelve were completed, but 11 of them required the developmental testers to intervene—and in some cases weaken the test rules to “less challenging” ones—to help the plane do things like acquire and identify the target so it could succeed in firing a weapon. Given these heavy interventions, DOT&E found that in its current configuration the combat effectiveness of the Marine Corps’ F-35Bs “will depend in part on the degree to which the enemy’s capabilities exceed the constraints of these narrow scenarios.” So the F-35 will win only if the enemy decides not to exceed the F-35’s limited capabilities.
The remaining three tests were pushed to later versions of the plane due to delays in implementing new software meant to fix mission system sensors and the data fusion problems. All of the deferred tests relate to the AIM-120 missile, the only weapon the F-35 can currently use against enemy planes.
Tests of the F-35’s ability to fire and drop the majority of its planned weapons in a combat-realistic operating environment won’t actually begin until the Block 3F configuration in 2021. Accomplishing those will require a total of 50 test events.
DOT&E believes these more complicated test events “cannot be accomplished within the remaining time planned by the Program Office to complete Block 3F flight test” in May 2017. This would require testing at triple the rate of what is being accomplished now. But the Block 3F tests will be much more complex and realistic than the current simpler engineering tests. It is unlikely more complex tests will be accomplished at the same rate as the simple testing, much less triple the rate. If, to make up the time, the program cancels many of these tests or defers them to the next Block as it has done in the past, “readiness for operational testing and employment in combat [would be] at serious risk.”
Pushing off tests only adds to what has become a compounding problem. The program currently has a 5 percent discovery rate for simpler developmental testing. This means that for every 100 tests, 5 new problems are discovered. These new discoveries then have to be fixed and tested again, which is a costly and time-consuming process. Even more troublesome, engineers are identifying problems faster than they can fix them. Inevitably, as testing continues and becomes more realistic, more and more problems will be identified, which will only draw out the process further. According to DOT&E, recent discoveries that require design changes, modifications, and regression testing (testing of the fixes) “include the ejection seat for safe separation, wing fuel tank over-pressurization, and the life-limitations of the F-35B bulkhead.” The F-35 is already years behind schedule. Issues like these are guaranteed to make the problem even worse.
The F-35 has had significant trouble with uncommanded “wing drop.” This means flaws in the aircraft’s aerodynamics under heavy maneuvering loads cause the aircraft to occasionally make sudden, uncommanded movements in the air. To fix this, the program made changes to the software, called the control law software, that translates the pilot’s commands into the actual movement of the plane’s flight surfaces. Those changes limit the maneuvers the pilot can command. Even with those changes, the plane is still experiencing excessive “buffeting”—intense shaking during certain fighting maneuvers because the airflow still separates from critical lifting surfaces under those manuevering conditions.
During one test flight of an F-35C, excessive buffeting “adversely affected performance in defensive maneuvering where precise control of bank angles and altitude must be maintained while the F-35C is in a defensive position and the pilot is monitoring an offensive aircraft.” In both defensive and offensive maneuvers, buffeting also made it difficult for the pilot to see the helmet-mounted heads-up display, which could significantly degrade pilot situational awareness and reduce chances of surviving the fight.
Buffeting and the reduced maneuverability caused by the associated control law software “fixes” featured prominently in the now famous example of the F-35 losing 17 dogfights to a 35-year-old, heavily laden F-16. In non-technical terms, the software fixes that provided a smoother ride for the pilot and lessened the uncommanded wing drop also limited his ability to turn hard enough to get away from an enemy plane on his tail. Similarly, when the F-35 finds itself on the enemy’s tail, the same fixes limit his ability to turn hard enough to keep up with an enemy plane trying to get away.
In an attempt to find a less compromising buffet fix, spoilers were fitted to test F-35s. These spoilers somewhat reduce the separation of airflow from the wing to lessen the shaking of the airplane during heavy maneuvering, and test pilots have reported some improvement in the buffeting as a result. But installing spoilers adds weight and increases drag, which only adds to problems the F-35 faces from using up almost all of its weight management safety margins. DOT&E questioned the net effect of the changes saying, “due to the transient nature of buffet, the operational significance may be low.”
Lt. Gen. Bogdan, the F-35 program executive officer, found himself hauled before a congressional subcommittee hearing in October 2015 after it emerged that he had grounded pilots weighing less than 136 pounds because mannequin tests showed that the ejection seat would kill them—and that no mannequin testing at all had been done for pilots weighing 137 to 244 pounds. The problem was a result of a number of faults in the seat design, exacerbated by the extra weight of the high-tech helmet.
This was far from the only F-35 safety issue engineers grappled with during the past year. For example, the F-35 Block 2B aircraft the Marine Corps claimed in July 2015 to be ready for combat had 27 serious safety deficiencies as of the end of October 2015. When DOT&E recognizes an issue, it is assigned to one of two categories based on severity and whether it threatens the safe operation of the plane. Category I is the most severe, being “those which may cause death, severe injury, or severe occupational illness; may cause loss or major damage to a weapon system; critically restrict the combat readiness capabilities of the using organization; or result in a production line stoppage.” The report lists a total of 91 current deficiencies, 27of which are Category I.
In a bit of good news, the Program Office was able to lift the restriction banning the F-35B from flying within 25 miles of known lightning strikes. The planes had been barred from doing so because the On-Board Inert Gas Generation System to add nitrogen and displace oxygen from the empty vapor spaces in the fuel tanks could not work fast enough to prevent an explosive mixture of fuel vapor and oxygen from collecting. The system has been fixed to the point now where the plane can fly in such conditions.
But it still can’t taxi or take off when there is lightning in the area. A problem with the software that controls the plane’s siphon tanks, which sit between the main tanks to keep the plane balanced as fuel is consumed, can cause too much pressure to build up and possibly cause a lightning-induced fire and explosion.
Although there are numerous hardware and structural issues remaining, problems with the software are much more likely to be the JSF Program’s undoing. Designers and engineers continue struggling through problems with the F-35’s approximately 8 million lines of onboard software code. Software remaining on the ground created even more headaches in 2015. The 24 million lines of complicated computer code running the maintenance and logistics program known called the Autonomic Logistics Information System (ALIS) has only lately garnered the serious attention required to make the system work. The DOT&E report provides details of a cascading problem of incremental software updates—or what Defense One dubbed a “terrifying bug list.” To cite just one of the many issues, the software can’t tell the difference between good parts and broken ones when it “incorrectly authorizes older/inappropriate replacement parts.”
The ALIS software went through four different versions in 2015: ALIS 1.0.3, ALIS 2.0.0, ALIS 2.0.1, and ALIS 188.8.131.52. While evaluating the software, personnel identified 2 Category I deficiencies and 56 Category II deficiencies in ALIS 1.0.3. One of the Category I deficiencies, for example, could prevent aircraft from taking off. The ALIS program keeps track of the maintenance status of planes in the fleet by generating a Health Reporting Code (HRC) for each plane. Should the software detect a technical problem in a plane, it creates a negative health report and depending on its severity, categorizes the plane as Non-Mission Capable. This effectively grounds the plane until the health report problem is diagnosed and repaired or, in the case of a falsely reported health problem, a supervisor overrides the system. These false positive health reports are not rare. Field reports say that 80 percent of ALIS-reported problems turn out to be false. This places a massive extra burden on the F-35’s already over-worked maintenance force.
Unfortunately, ALIS makes supervisor overrides to prevent grounding very difficult. Two modules in the ALIS system prevent overrides, even after a recent ALIS software update. The aircraft computer generates the health reports, which are downloaded into an ALIS module called the Computerized Maintenance Management System (CMMS). Another module, the Squadron Health Management Module, makes a mission-ready determination based on Mission Essential Function List. One module might declare an aircraft mission ready while the other asserts the opposite. In those cases a maintenance supervisor who has determined that the aircraft is ready should have the ability to override the system in order to clear the plane to fly. Unfortunately a software problem in the Squadron module prevented this from happening.
The test team noted this problem remained in ALIS 2.0.0 (in addition to finding five more Category I deficiencies) and still remained in ALIS 2.0.1. Developers finally fixed the problem with a “patch” in ALIS 184.108.40.206, yet five more major problems were discovered.
The report also notes that all versions of the ALIS software have problems with data quality and integrity. This is particularly true with the system’s Electronic Equipment Logbooks (EELs), a system for tracking aircraft parts. This system frequently fails to create accurate entries or to transfer data properly, forcing maintenance crews to waste time with manual workarounds. According to the report: “Without accurate EELs data, ALIS can improperly ground an aircraft or permit an aircraft to fly when it should not.”
History and experience suggests the problems with ALIS are only beginning. The more complex and lengthy a software program is, the better the chance coding errors will plague the system. According to one IT consultant, even in well-written programs, developers find bugs at a rate of 1 per every 1,000 lines of code. In fixing the one problem, software patches tend to introduce new bugs and security vulnerabilities at a rate of 10 to 15 percent. The DOT&E report certainly appears to confirm this in the case of ALIS.
Nearly all of the promised capabilities of the F-35 rely on its sophisticated network of computer-based systems, both on the ground and in the plane itself. The sensors to locate and identify enemy targets, guidance systems to direct missiles and bombs, diagnostic tools to isolate defective parts and order spares, the pilot’s helmet-mounted display, and even the mission order packages all operate on computers and complicated software. As has been repeatedly proven over the years, systems like these are tempting targets for hackers. Pentagon officials have already acknowledged the F-35 program suffered a major breach when a foreign power, presumably China, hacked into an unclassified F-35 contractor computer network and stole massive technical data files.
But despite these risks, the Joint Program Office has refused to subject the program to the kind of cyber testing necessary to identify and fix vulnerabilities. As we previously reported, the JOTT created a two-part test plan to evaluate the program. The first, an internal assessment to comb through the system’s designs to identify potential problems, was only partially completed on isolated modules at Edwards Air Force Base. Even these limited tests revealed “significant deficiencies,” although the DOT&E report did not provide any details as to their nature.
The second crucial phase of testing, which unleashes DoD “Red teams” of hackers to break into the system, did not happen at all. General Bogdan refused to grant permission for the Red Team tests “due to insufficient understanding of risks posed to the operational ALIS systems by cybersecurity testing.” Put in less obscure terms, the F-35 program management cancelled tests of a combat-critical computer system because they thought the tests might break the computer system.
The reason testing takes place is to ensure the programs and systems the services are buying work properly in peacetime and war. The program office validated the strong need for F-35 cyber testing in the reasoning they gave for cancelling it. The computer glitch that allows ALIS to ground an aircraft would be an obvious target for an enemy cyber warrior. Even more tempting would be stealing mission order packages or planting false ones. If peacetime cyber testing can damage the ALIS system, what could a determined enemy do?
All of the time and money expended on the F-35 will have been for naught if the plane can’t get off the ground when it is most needed. Unless the program improves dramatically in basic availability, nearly half of the F-35s in the fleet will not be able to fly at any given time due to a variety of persistent maintenance issues. Maintenance crews have had so much trouble keeping the aircraft flight-worthy that most planes fly less than twice in a typical work week.
During 2015, 10-20 percent of all F-35’s in service were undergoing major overhauls, according to the report. Of those that remained, only “half were available to fly all missions of even a limited capability set.” The program had set a goal of 60 percent availability to fly for 2015, but the entire fleet only averaged 51 percent. This actually represented a marked improvement“over the 37 percent availability reported in both of the previous two DOT&E Annual Reports from FY13 and FY14.” However, it still falls far below the 80 percent availability rate generally considered minimally adequate for any military aircraft on a real combat deployment.
The DOT&E report reviews a number of maintenance metrics, but those metrics are suspect because they are stored by the contractor in an unsecured database and, according to the report, is neither current nor validated by government oversight. Evaluators are actually prevented from accessing the database from government networks because Lockheed Martin’s database does not meet U.S. Cyber Command cybersecurity standards. The evaluators have received hard copies of some of the data but have been unable to review and validate all of the contractor maintenance records including availability rates and reliability numbers.
While the program’s specific maintenance numbers remain obscure, the sortie rate is not. The test aircraft at Edwards Air Force Base were only able to fly once every 5 days (6 flights per month). Operational units at other bases reported similar numbers: Luke AFB F-35s averaged 1 flight every 5 days, and F-35s at Nellis AFB averaged 1 flight every 4.75 days.
The report lists a few of the “High Driver Components Affecting Low Availability and Reliability,” or the most common broken parts affecting the fleet in general. Computer components on all variants failed at a high rate, as did fuel pumps and main landing gear tires. Crews also had to work hard to fix problems with the plane’s stealth coatings. Low Observable Maintenance, or fixing stealth components, is time consuming in part because the necessary skin panels, sealants, and paints to make the plane harder to track with radar are delicate, have long cure times, and are potentially highly toxic. “From July 2014 to June 2015, program records show that maintenance on ‘attaching hardware,’ such as nutplates and heat blankets, absorbed approximately 20 percent of all unscheduled maintenance time, while low observable repairs accounted for 15 percent,” according to the report.
In a combat situation or for large training exercises, units operate at what is known as a surge rate and, to meet short-term demands, must generate more flights per day than the sustained rate. To do so, crews work overtime beforehand to ensure a maximum number of aircraft are prepared to fly and then afterwards to repair the extra maintenance issues put off during the surge period.
The Marine Corps surged F-35 operations to support Operation Steel Knight, a large-scale air-ground exercise at Twentynine Palms, California, in December 2015. They planned to use 8 F-35Bs to fly close air support missions for the ground units over the course of 8 days. The schedule called for each aircraft to fly approximately 1 sortie per day. But even under short-term surge conditions, the squadron could only manage 1 sortie every 2.3 days. According to the report, “while deployed, and in support of the exercise, the Marine Corps flew approximately 46 percent of the planned sorties (28 sorties flown versus 61 sorties planned), not including the deployment, redeployment, and local familiarization sorties.”
Operation Steel Knight is an annual event for the Marine Corps. Detailed planning for it begins at least six months before the first units move out to the field. Maintenance crews had months to prepare the necessary aircraft to support this exercise and they still barely managed to get the planes to fly once every 3 days. A future enemy will likely not be so considerate as to provide advanced notice.
The only way to test many of the F-35’s capabilities is in a virtual simulated environment because the test ranges cannot accurately replicate the full spectrum and quantity of threats the jets would confront. Contractor engineers have been tasked since 2001 with creating a testing facility called the Verification Simulator (VSim). It was intended to be an ultra-realistic, thoroughly test-validated “man-in-the-loop, mission systems software in-the-loop simulation developed to meet the operational test requirements for Block 3F IOT&E.” The final decision Congress makes to go into full-rate production will be based on tests conducted in facility like this. A similar system, the Air Combat Simulation (ACS), was used by the F-22 program to fly scenarios not possible in open-air range tests using realistic threat numbers and tactics. According to the report, the facility fell hopelessly behind and has now been reassigned to a government agency.
For over five years, DOT&E has raised concerns about the failure of the project and its dire consequences for completing adequate operational testing. In 2010 DOT&E, faulting the JSF Program Office for assigning low priority to and shortchanging the project, bluntly “identified funding shortfalls for the Verification Simulation (VSIM) to meet OT&E needs, primarily in the battlespace environment, and provided data for an independent cost assessment leading to inclusion of VSIM costs in the program baseline.”
Following the 2010 Nunn-McCurdy restructuring of the JSF program, $250 million in funding was added to the F-35 budget for the Verification Simulation Facility. Despite the potential for conflicts of interest, the program office rejected a plan for the government to build the simulator in 2011 and decided to leave the contract with Lockheed Martin, but then, in August 2015, the Verification Simulation project was transferred to Naval Air Systems Command (NAVAIR) because so little progress had been made, Fifteen years after the project began, it is now beginning all over again from scratch.
To create and validate this high-fidelity virtual world suitable for combat test missions, the simulation designers will need to conduct many actual F-35 flights to gather onboard data on maneuvering performance, handling qualities, flight controls, radar, infrared imagery, weapons trajectories, and homing behavior in the presence of representative terrain and realistic ground and air threats. This is intended to be the basis for digitally recreating “the F-35 and other supporting aircraft, and models of airborne and ground-based threats.” These models are to be combined with information about the projected threats to build a full Battlespace Environment capable of realistically simulating large arrays of friendly and enemy forces to test the F-35’s combat effectiveness in the complexity of real combat.
As an example of the difficulty and scope of the needed validation effort, a test pilot will fly a mission over a range set up with multiple enemy radar and missile systems. The real F-35’s sensors, electronic warfare system, and intelligence links detect this threat and respond accordingly, providing pilot warnings, signal jamming, defense suppression missile firings, or any number of other responses. All the data, massive amounts from all the instruments, from the real-world test flight is gathered and then compared against the data from the Verification Simulation’s re-creation of exactly the same flight test scenario. When such measured outcome comparisons show reasonably similar behavior over many test flight scenarios, then the Verification Simulation can be declared valid. This process takes years, and the fact that it hasn’t been diligently pursued since its inception 15 years ago puts the entire effort even further behind.
The actual simulation facility was intended to have four high fidelity F-35 cockpits and eight additional threat and friendly aircraft control stations to allow real people to “fly” complex missions with and against the F-35s. This was to conduct tests of large multi-ship flights of F-35s against dense air and ground threats. These are the hardest, most expensive, but most important operational tests to conduct with real airplanes and threat simulators.
The Verification Simulation facility was also intended to test the networking of onboard and offboard sensor and intelligence data between all F-35s in a formation. This networking capability remains one of the biggest selling points of the plane. Called data fusion, it is supposed to create an identical operating picture for all the pilots during a mission. So far, the program has encountered major deficiencies in data fusion in even the most basic engineering flight tests. Specifically, the onboard computers have been unable to usefully merge the target data from fourF-35s flying in an area free of enemy interference. Using the Verification Simulation facility in more realistic and complicated combat test scenarios would likely uncover even more problems. If the F-35 program proves unable to deliver this capability, the entire reasoning behind the program would be questioned.
DOT&E reserved some of its harshest criticism for failures in the Verification Simulation project. “Due to inadequate leadership and management on the part of both the Program Office and the contractor, the program has failed to develop and deliver an adequate Verification Simulation (VSim) for use by either the developmental test team or the JSF Operational Test Team (JOTT), as has been planned for the past eight years and is required in the approved TEMP.”
DOT&E does not have much confidence in NAVAIR’s ability to construct the necessary facility to fully test the F-35 in time to meet the current test schedule. “It is also clear that both NAVAIR and the Program Office significantly underestimated the scope of work, the cost, and the time required to replace Lockheed Martin’s proprietary BSE (Battle Space Environment) with the JSE (Joint Simulation Environment) while integrating and validating the required high-fidelity models for the F-35, threats, friendly forces, and other elements of the combat environment.”
Without a validated simulation facility the F-35 program would have to conduct “a significant number of additional open-air flights during IOT&E, in addition to those previously planned” in order to complete testing on time. Since the plane already can’t fly often enough for the current developmental testing schedule, expecting to be able to stuff in the necessary additional flights is unreasonable.
At best the Program Office merely dropped the ball in failing to devote the proper amount of effort to establish a needed facility. Congress should audit what happened in the VSim program to determine why it failed and whether taxpayers deserve a refund. The program office’s failure increases the risk that shortcomings with the F-35 program may only be revealed in actual combat. This would likely result in failed missions and needless casualties.
The DOT&E report also provides further proof that the Initial Operational Capability (IOC) declaration by the Marine Corps last summer was nothing more than a public relations stunt and that the Air Force’s planned declaration later this year will be as well. Then-Marine Corps Commandant Gen. Joe Dunford (now Chairman of the Joint Chiefs of Staff) announced on July 31, 2015, that the Marine Fighter Attack Squadron 121 at Yuma, Arizona, “has ten aircraft in the Block 2B configuration with the requisite performance envelope and weapons clearances, to include the training, sustainment capabilities, and infrastructure to deploy to an austere site or a ship.” In other words, the Marine Corps claimed to have 10 F-35s ready for combat and enough spare parts and maintenance personnel to support the squadron.
But DOT&E found that significant combat deficiencies remain. “If used in combat, the Block 2B F-35 will need support from command and control elements to avoid threats, assist in target acquisition, and control weapons employment for the limited weapons carriage available (i.e., two bombs, two air-to-air missiles),” wrote Dr. Gilmore. The report also states, “If in an opposed combat scenario, the F-35 Block 2B aircraft would need to avoid threat engagement and would require augmentation by other friendly forces.” This means the F-35Bs the Marine Corps said are ready for combat would need to run away from enemy planes while other aircraft would be needed to come to their rescue.
Air Force officials have repeatedly stated their plans to declare Block 3i of the F-35A—the conventional take-off model—combat ready in August (with a December fail-safe date), as scheduled. Block 3i configuration has a newer computer but the same extremely limited weapons and combat capabilities as the 2B. On the current schedule, the Air Force will declare initial combat capability with planes that, like the Marines’ variant, will have to run from enemy fighters, need other airplanes to help find targets and avoid threats, and carry only two air-to-air and air-to-ground weapons.
Testing has actually revealed that the newer hardware and software used by the Air Force is sometimes in worse shape that the earlier Block 2B used by the Marines. This is especially troubling because officials had written the Block 3i testing plan simply “to confirm Block 3i had equivalent capabilities to those demonstrated in Block 2A (for 3iR1) and Block 2B.” The testing office originally planned for 514 baseline testing points. During the baseline testing, another 364 additional “discovery” testing points were identified. This means that during testing, 364 additional tests had to be added to try to fix newly discovered problems in a system that was already supposed to work and only had added a new computer. For example, DOT&E reported the unacceptable “instability” (that is, frequent crashing) of the Block 2B computer-based radar. In fact, Block 3i radar performance was found to be “less stable” than Block 2B. The 3i radar now crashes 7.5 times more often than the earlier version.
The F-35A will be hampered with limitations on both basic flight and weapons employment. The same problem preventing F-35s from taking off in a lightning storm also prevents them from performing hard maneuvers with full fuel tanks. Fully fueled F-35’s are limited to only 3 g’s because harder maneuvering could increase the pressure in the siphon tanks beyond their limits. The plane is also limited from opening its weapons doors to fire at speeds above Mach 1.2 due to concerns about structural vibrations called “flutter.” This is less than the plane’s maximum allowable speed of Mach 1.6. Since the F-35 was sold as a supersonic fighter, this restriction negates one of the major capabilities used to justify the massive bill to the American people.
In a congressionally mandated 2013 report, the Department of Defense set the dates and criteria for IOC. In the case of the Air Force, “F-35A IOC shall be declared when Airmen are trained, manned and equipped to conduct basic CAS, Interdiction, and limited SEAD/DEAD operations in a contested environment.” The Air Force set its target IOC date as August 2016 with December 2016 as a backup. The report also states, “Should capability delivery experience additional changes, this estimate will be revised appropriately.”
As is clearly evident in the DOT&E report, the criteria necessary for the Air Force to declare IOC have yet to be met. The aircraft will have little, if any real combat capability for years to come. And with as much trouble as the services have had keeping their planes flightworthy, it is nearly impossible for all the pilots to have acquired enough real flying hours to develop the combat skills they need. Despite these issues the Air Force is widely expected to make its declaration on time in August. Like the Marine Corps’ declaration last year, it will be nothing more than a PR stunt meant to keep money flowing into the program.
As part of the efforts to reform how the Pentagon buys equipment, Under Secretary of Defense for Acquisition, Technology, and Logistics Frank Kendall has urged the Pentagon “fly before you buy.” The F-35 program has done the opposite. Current purchase plans would see the services with approximately 340 F-35s by the end of the next fiscal year, long before IOT&E is complete. Instead, the F-35 program has experienced an unprecedented level of concurrency, approving increasing levels of production years before development and testing can possibly be completed.
The GAO estimates concurrency in the F-35 program will cost $1.7 billion to “rework and retrofit aircraft with design changes needed as a result of test discoveries.” As planes continue to come off the production line long before testing has uncovered all the design defects, much less their fixes, that figure will dramatically increase.
The level of concurrency in the F-35 program causes it “to expend resources to send aircraft for major re-work, often multiple times, to keep up with the aircraft design as it progresses.” (Emphasis added) Some retrofits are a normal part of the acquisition process. But the level of production and rate of newly emerging design failures mean there are an unprecedented number of planes that must be altered at significant expense. For example, by the end of 2017 the program will have delivered nearly 200 aircraft that almost certainly will not be in the 3F configuration necessary for IOT&E.
There is a very real danger some of the problems can’t be fixed within affordable budgets. During static strength and fatigue testing there have been large numbers of demonstrated structural flaws, including cracking and metal fatigue in the wing structure, fuselage bulkheads, and almost every door on the airplane. DOT&E cautions the services may be stuck with numerous left-behind aircraft they can’t afford to upgrade: “these modifications may be unaffordable for the Services as they consider the cost of upgrading these early lots of aircraft while the program continues to increase production rates in a fiscally-constrained environment.” These concurrency orphans would likely serve as little more than costly sources of spare parts or un-representative test beds.
The cost to implement retrofits and the purchase price of planes made obsolete because they never are fixed add up to the program’s “concurrency tax.” With several years of development and testing still to come, the amount of this tax will continue to spiral ever upwards.
JSF Program officials both inside the government and Lockheed Martin have repeatedly expressed their desire to move beyond low rate initial production. They want Congress to authorize a block buy for 465 planes—with commensurate large pre-payment—for the United States and foreign military partners beginning in 2018. General Bogdan claims such a move would save “billions of dollars.” The DOT&E report not only pokes holes in the cost-saving claims, but more importantly questions the legality of such a commitment. It is perhaps telling that officials are seeking a block buy at this point rather than a multi-year purchase contract.
Federal law allows multiple year contracts to purchase government property so long as certain criteria have been met. Congress typically authorizes most weapons buying programs on a year-by-year basis to ensure proper oversight of the program and to maintain incentives for the contractor to satisfactorily perform. According to Title 10 U.S.C., Section 2306b, for a program to be eligible for multiyear procurement, the contract must promote national security, should result in substantial savings, have little chance of being reduced, and have a stable design. The F-35 seems to be failing at least two of the first three criteria and is most certainly is failing the fourth.
As the DOT&E report shows, the operational testing that needs to take place in order for an informed final production decision will not be completed until 2021.
Multi-year contracts afford some protections to the taxpayers. But the program office is proposing a block buy, which provides significantly fewer protections for taxpayers. As a Congressional Research Service report points out, block buy contract savings can be lower than those promised under multiyear procurement, and are not governed by any precautionary statutory requirements.
The JSF Program has already been in development for more than twenty years. The plane is still years away from being capable of providing any real contribution to the national defense if, in fact, it ever will be. The issues raised with this program are important for everyone, citizens and decision-makers alike to understand. There is already discussion in the halls of the Capitol and the corridors of the Pentagon about the next fighter plane program beyond the F-35. Unless everyone learns from their mistakes with this program, history will be repeated. The United States can ill-afford another $1.4 trillion mistake that will do more to harm our national security than it does to secure it.
The DOT&E report makes perfectly clear that any further F-35 production at this point is unwise. The plane has yet to prove itself capable of performing even the basic combat tasks used to originally sell the program to the American people. Congress should scrutinize carefully any further production proposal, as only the contractors will benefit from turning out airplanes that can’t fight and that will carry a crushing retrofit bill. The rest of us—particularly those who fight our wars—will be left to bear the cost.
The Center for Defense Information at POGO aims to secure far more effective and ethical military forces at significantly lower cost.