There is an awful lot of software on this program. It scares the heck out of me.”

That was U.S. Air Force Lt. Gen. Christopher Bogdan's candid reaction to the F-35's dependence on yet-to-be-delivered software before he assumed the top spot overseeing the $390 billion program in December 2012. At that time, he was serving as the program's deputy and was just getting his head around the massive software effort behind various capability releases that would underpin the operational readiness of the three F-35 variants.

The Pentagon's chief tester, Michael Gilmore, says Bogdan should be scared.

Gilmore predicts in his annual operational test and evaluation report, which went to Congress Jan. 27, that software could drive an up-to-13-month delay in the readiness of the 2B capability release the Marine Corps needs for initial operational capability (IOC), planned for July 2015.

This could be a major hurdle for the multinational program, which Pentagon leaders have said will receive no more money. The program has been rife with restructurings and multibillion overruns ever since development started with prime contractor Lockheed Martin in 2001.

Aside from political pressure to stay on schedule—to retain support from nine international partners and two foreign military sales customers—the Marine Corps has plans to begin an operational utility evaluation of the F-35B with the 2B software in early 2015, in advance of its IOC. This depends on the availability and reliability of the 2B software.

Bogdan says: “There were no surprises in the report; all of the issues mentioned were well-known to us.” But it does give the public a peek into a detailed assessment of progress for this complex system; often what is known to the program office is not necessarily clear to the public or Congress, which is where the funding comes from. The testing office is often seen as conservative in its predictions, while program offices are typically more aggressive. Findings in past reports have helped officials tackle such problems as lightning vulnerability and an onboard oxygen-generating system.

Because the software releases are inter-layered, a delay to 2B would affect Block 3i (which includes some new processing hardware and with which the Air Force intends to declare IOC in December 2016), and Block 3F, which allows for far more robust combat capability through the use of external weapons, a wider variety of weapons and a more complete engagement capability.

Testing for the Block 2B capability—which encompasses use of the 1,000-lb. Joint Direct Attack Munition, GBU-12 Laser-Guided Bomb and AIM-120 Advanced Medium-Range Air-to-Air Missile—began several months late due to the additional time needed for 2A, an earlier block. Block 2A allowed for basic flight training to take place while developers continued flight-test work for later iterations. “A considerable amount of testing was necessarily devoted to completing development of prior-block capabilities, attempting to complete fixes to known problems, and regression testing of new versions of software,” Gilmore's report states. “This creates significant pressure on development and flight-test of the remaining increments of the Block 2B, with approximately 12 months remaining on the program timeline before final preparations are planned to begin for an operational utility evaluation of the combat effectiveness and suitability of Block 2B.”

Because testers found that an initial version of the 2A software, 2AS3, was “deficient in providing the necessary capabilities for unmonitored flight operations under night and instrument meteorological conditions,” the F-35 team developed software version 2AS3.1 before they could move on to 2B.

Problems addressed in the 2AS3.1 version included anomalies with the electro-optical targeting system (EOTS), aircraft communications, electronic interfaces for the pilot, and the caution, advisory and warning system. “The software was intended to provide more mission-systems capability, but poor sensor performance and stability, excessive nuisance warnings, and the disproportionate pilot workload required for workarounds and system resets made the software of limited utility for training,” Gilmore says.

A jump of 40% above what was allocated for mission-system testing—including a series of trials specifically addressing shortcomings in the helmet-mounted display system—contributed to Gilmore's grim prediction. He says, a 40% rate of growth in test requirements, coupled with the team's pace at handling the trials, means “Block 2B developmental testing will complete about 13 months later, in November 2015, and delay the associated fleet release to July 2016.”

The plan calls for 2B testing to be completed by the end of 2014.

Using the 2013 test point growth rate as a metric for performance in 2014 is not accurate, however, according to Lockheed Martin spokeswoman Laura Siebert. “Twenty-two percent of last year's test-point growth was directly attributable to dedicated Helmet testing, [which] will not be required in 2014.”

However, the team is still investigating why the EOTS is failing to meet its expected target-recognition ranges; pilots will be using EOTS in combat to designate and attack targets. But it “exhibits track instability in portions of its field-of-view, and has large line-of-sight angle and azimuth errors when computing target locations,” Gilmore says.

The distributed aperture system (DAS) (which provides 360-deg. situational awareness through six externally mounted cameras) “has displayed a high false-alarm rate for missile detections during [single-ship] and formation flare testing. The inability of the DAS to distinguish between flares and threat missiles makes the warning system ineffective.”

Also, the onboard navigation tool has shown an inability to operate effectively without GPS updates, hindering a pilot's ability to target weapons in a GPS-denied environment.

The program was designed to rely on sequential software releases. But with the next iteration, 3i, incorporating a new core processor, three mission-system aircraft (one of each variant) are expected to be pulled from the test force for the modification to prepare for future trials for that block. Already an attempt for early 3i testing on an F-35C, which was temporarily modified with the hardware, failed because the software did not load properly. Gilmore predicts the transition to 3i will add pressure to the 2B test schedule. “All program plans and schedules for the subsequent blocks of mission systems software (Blocks 3i and 3F) depend on [2B completion] so the development laboratories and test venues can be converted and devoted to testing the Block 3 hardware configuration.”

Siebert notes that there will be a test aircraft added to the total fleet for mission system work, although three will be dedicated to 3i work.

Despite the potential delay, the Marine Corps is still planning for an IOC as early as July 2015. “We are not changing our plan,” says spokesman Capt. Richard Ulsh. Lt. Gen. Robert Schmidle, deputy commandant for aviation, says the F-35, even with 2B software, offers more capability than the current Hornet and Harrier fleets.

Bogdan's office and Lockheed Martin remain confident they will deliver the 2B software to support a July 2015 IOC for the Marines. The program office has shown no signs of studying whether shifting IOC focus to 3i, which is slated for operations roughly a year after 2B, would relieve schedule pressure.

“Although the report is factually accurate, it does not fully highlight the F-35 enterprise's efforts to address and resolve the known technical and program-related challenges,” Bogdan says. “Software continues [as] our No. 1 technical risk and we have instituted disciplined systems engineering processes to address the complexity of writing, testing and integrating software.”

Company officials predict they will finish 2B flight-testing this year as planned. Siebert says that 300 mission system flights are slated for 2B work this year, the same number of mission system flights executed last year. Lockheed is “fundamentally on plan” with the schedule, based on the 2010 restructuring to the program, she says. F-35 Joint Program Office spokesman Joe DellaVedova acknowledges the 2B work is four months behind plan, but still within the allowed margins.

Gilmore's report also lists reliability and availability metrics for the young F-35 fleet. Though small, the force at Eglin AFB, Fla., and Marine Corps Air Station Yuma, Ariz., is growing. With this information, the Pentagon can begin gathering statistics on the health of the fleet. Thus far, it is falling short of expectations for both availability—mission-ready aircraft when needed—and reliability (see chart).

On average, the fleet was available 37% of the time, far below the threshold goal of 50% and the objective of 75% (which is included in performance-based logistics agreements). This is related to several factors, including: slow response from Lockheed Martin on some “action requests” for input on maintenance; inadequate supply; and from maintainers taking longer than predicted to handle tasks. The program office is aware of these issues and has established readiness cells to tackle the most challenging factors contributing to the lackluster performance. An air-system readiness cell is tracking the points where timeliness is compromised, says DellaVedova. He notes that readiness factors are woven into the incentive-fee structure of the annual lot production contracts for Lockheed.

Gilmore is also concerned about subpar performance of the automated information logistics system (ALIS), a sweeping database used for mission planning, maintenance and supply functions. A fully capable ALIS, which is not yet available, will allow program officials to conduct predictive maintenance, allowing trend data to highlight when parts would be needed in advance of failures. The prognostics, however, are not yet functioning, partly because ALIS immaturity is forcing operators to manually input data that will eventually be automated. DellaVedova says the predictive maintenance capability will come when the fleet is larger and can provide a more comprehensive pool of data.

The program is roughly halfway through its test phase; development is slated to be finished in 2016. “We recognize risks still exist in the program, but they are understood and manageable,” Bogdan says. “The F-35 is currently in its developmental phase. This is the time . . . to discover issues through testing so we can implement solutions.”

With Bill Sweetman in Washington.

Read the Pentagon chief tester's fiscal 2013 report on the F-35 development process at

F-35 Components Contributing to Low Availability & Reliability
Specific to Variant Common to All Variants
F-35A • Data transfer cartridge• Position/strobe light lens assembly • 270-volt DC battery• Fiber channel switch• Avionics processor• Power and thermal management system• Landing gear and tire assembly• Display management computer/helmet• On-board oxygen-generating system• Crew escape and safety system• 80-kw inverter/converter/controller
F-35B • Upper lift fan door actuator• Main landing gear wheel/tire assembly
Source: U.S. Defense Department
Operational F-35 Availability & Reliability (hours)
Variant Aircraft Availability Mean Time for Critical Failure Fixes Mean Time for Repairs
F-35A Goal: 20Actual: 4.5 Goal: 4Actual: 12.1 Goal: 2.5Actual 9.2
F-35B Goal: 12Actual: 3 Goal: 4.5Actual: 15.5 Goal: 3.0Actual: 8.9
F-35C Goal: 14Actual 2.7 Goal: 4Actual: 9.6 Goal: 2.5Actual: 7.7
Source: U.S. Defense Department
Fiscal 2013 F-35 Testing Fast Facts
Average operational F-35 availability rateGoal: 50% threshold and 75% objectiveActual: 37% Average unavailable aircraft due to maintenanceGoal: 6%Actual: 35%
Average unavailable aircraft due to parts/supplyGoal: 20%Actual: 27% Test Points Goal: 7,180Actual: 5,464Test SortiesGoal: 993Actual: 985
Source: U.S. Defense Department