JSF - concurrent development and manufacture: is this really a problem?

I was struck recently by an online story on the Flight Global web site, from which I reproduce the following three paragraphs:

“The [US] Department of Defense agreed with a Government Accountability Office report that it should resolve the Lockheed Martin F-35’s critical deficiencies before requesting funds to pay for full-rate production of the aircraft.

“Waiting for the resolution of critical deficiencies in the F-35 Lightning II could delay the aircraft’s full-rate production, which the DoD had planned to begin in 2021.

“The department had planned to defer resolving some critical deficiencies found in testing until after its full-rate production decision in October 2019, according to a GAO report released on 5 June. However, the report cited concerns that fixing deficiencies after starting full-rate production could eventually create additional costs tothe government.”

Yes, I can understand the logic behind the GAO recommendation, but it’s this very linear approach to multi-path innovation that creates (or compounds) many of the project delays that bedevil modern, complex acquisition projects such as the F-35 and, before it, the F-22 and B-2.

If you were going to retire all the risk on every aspect of the project before approving funding for the next stage or for Full-Rate Production you’d probably never complete the project. At the very least, you’d be looking at one hell of a slow schedule. 

Just to make it clear, I don’t have a problem with concurrent development and manufacture. In fact, I would suggest that the practice be endorsed more frequently by the Joint Program Office (and CASG here in Australia). However, this would only work if expectations were subtly modified. Read on. 

Consider: the F-35 program embodies all-new technology in multiple areas concurrently: airframe manufacture; mass-production of stealth features; helmet-mounted display; RF and IR sensor suites; flight control software; engine and vertical lift system; in-service support via the ALIS; and so on. 

Yes, you could probably prove many of these individual technologies (perhaps not the airframe, stealth mass production and vertical lift system, admittedly) in a bunch of separate risk-mitigation projects using existing platforms, or even just design and build a bunch of different aircraft that embody whatever selection of these new technologies is useful to their specific role. But I note that the three US customers for the F-35 (USAF, USN and USMC) chose not to do that. And none of the export customers had the resources to even contemplate doing that by themselves or in a non-US consortium, except for the Eurofighter consortium members – and the Eurofighter program predates the JSF by a decade.

Therefore, the decision to develop and build an all-new 5^thgeneration aircraft that embodied all of these technologies was probably inevitable – as was the subsequent decision to pursue concurrent development and manufacture of the F-35. If the JSF partner nations hadn’t agreed to do that we’d probably still be waiting for Low-Rate Initial Production (LRIP) to begin. 

Instead, we have more than 265 aircraft already delivered; the F-35A and F-35B are already in front line service with the US Air Force, Israel Defence Force and US Marines, respectively; and the F-35A has already been to war against the modern air defence systems the Russians have sold to their ‘friends’ in Syria. And the F-35 hasn’t even entered Full-Rate Production (FRP) yet. 

Think back 80 years. In 1938 the Royal Air Force was just starting to re-equip its fighter squadrons with their first monoplane fighters, the Hawker Hurricane and the Vickers Supermarine Spitfire. The Hurricane was the service’s first 300mph fighter, but it was largely fabric-covered, had the most basic version of the Rolls-Royce Merlin engine developing barely 1,000hp, and it was propelled by a fixed, two-bladed airscrew. The Merlin was still a developmental power unit and the fighter’s new reflector gun sight and eight-gun armament were also new and experimental. The Spitfire was a much more modern design but still had the same engine, twin-bladed airscrew, armament and gun sight. Both aircraft represented a revolutionary advance in air combat technology for the RAF.

But the early model Hurricanes and Spitfires, advanced as they seemed for their time, were only at the start of their development paths. Two years later, in mid-1940, just as the Battle of Britain was getting under way, they both had a much more powerful engine and a Rotol three-bladed, constant speed airscrew that transformed their take-off and altitude performance. The Hurricane also now had all-metal wings that reduced overall weight by 80lb while increasing maximum speed by 80mph. These were very significant improvements for their time. 

Importantly, the engine and airscrew upgrades were implemented on both aircraft as fleet-wide retrofits. The same went for the Hurricane’s metal wing. Was this expensive? Undoubtedly. But these retrofits were essential – during the early months of World War 2 the Hurricanes’ old fabric-covered wings and two-bladed airscrews put them at a huge disadvantage.

The point is that, firstly, the RAF had the aircraft in service to be retrofitted: if production had been delayed to allow testing of these and other important enhancements to be completed the RAF would have been woefully under-equipped and quickly overwhelmed. Instead, production continued and existing aircraft were retrofitted once in service; the sheer numbers mattered. Secondly, as a consequence, the RAF had two years before the start of World War 2 to train pilots and controllers and develop tactics and procedures to use the Hurricane and Spitfire effectively. Importantly, also, the RAF learned in peace time how to operate, sustain and repair its new generation of aircraft (including solving critical gun freezing problems at high altitude), not something you want to learn in a hurry, ‘on the job’. 

Fighter Command learned how to fight as a single complex, integrated organism of which the fighters were a critical part, though by no means the only one. Victory in the Battle of Britain was a collective achievement that couldn’t have happened if production of the Hurricane and Spitfire had been delayed. 

So how does this relate to the F-35? Early versions of the Hurricane and Spitfire were barely good enough for the job, but operational experience showed where investment was most needed in performance and combat capability improvements. Many of those improvements, such as reflector gunsights on the early Hurricanes and metal ailerons on the Spitfires, were retrofitted. 

The same applies to the F-35: if we accept that the aircraft will need to be developed through its service life to remain both relevant and superior, then we need to understand that a stated level of capability is merely a waypoint on a journey, not a final destination. It’s more important to have a design and platform that can be enhanced continuously and a mechanism for capturing constantly the creativity, technical smarts and operational insights that will guide each subsequent stage on the journey. In the meantime a less than ‘perfect’ aircraft can still provide sterling operational service and will provide invaluable training for both pilots and air forces.

Here’s a final thought from the 1940s: the Spitfire remained in production for a decade, through 24 marks; it saw the introduction of a new engine, Rolls-Royce’s Griffon which had nearly twice the power of the early Merlins, and finished with twice the take-off weight and twice the range of the initial version. Almost every aircraft since then that has had an extended production and service life has enjoyed similar growth in operational capability. The final capability in each case has been unforeseeable to the original designer of the aircraft in question, but usually resulted in the aircraft themselves becoming legends. Think F-86 Sabre, F-4 Phantom, F-15, F-16, F/A-18, Hunter, Harrier, Canberra, Su-27, MiG-21, Tornado strike variant and Mirage III and 2000. 

Having something that’s good enough today can be more valuable than having something that’s perfect tomorrow. Firstly, you need to survive today in order to have any hope of making it as far as tomorrow; secondly, we can’t foretell the future, so nobody really understands what ‘perfection’ will mean tomorrow in any case. So we need to design for what we can see and credibly predict; once the aircraft (or tank, or submarine) is in service we can predict the next developmental stage with much greater confidence. We can build an upgraded design to meet the new challenge, but we should also incorporate as many as possible of these changes on existing aircraft (or tanks, or submarines). 

If that means concurrent development and manufacture, then so be it. If the alternative is an interminable wait for a mythically perfect machine which enters service too late to be relevant, that’s unacceptable. 

ENDS