Sunday, February 16, 2014

Concurrency and the F-35: A CBS 60 Minutes (Re-Run) Viewers Guide

 Courtesy of yesterday, I got a late head’s up on one of the Anti-JSF (politico, journolista, and ‘reformer’) memes that 60 Minutes is going pick up in their JSF segment tonight: “There’s been too much ‘Concurrency’ on the F-35 program”. How far 60 Minutes will run with the meme I don’t know, but I thought I’d highlight here just how the F-35 is ‘different’ from most (and ALL Air Force) post-WW2 predecessor programs, by highlighting the one aspect where it is viewed differently from all its predecessors.
While I’ve compiled prior examinations before on this topic, exposing the Concurrency Canard for what it was, and further reemphasized same when more supporting information became available, I think perhaps a review of the actual historical record should further drive the point home that ‘Concurrency’ as it applies to the F-35 is merely a smokescreen used to distract the ‘ahistorical’ among us.

Using “Post-World War II Fighters: 1945-1973” (contents, apparently verbatim, are also found here) for the ‘old’ systems, we will find that to varying degrees, fighter aircraft have always been fielded with originally planned capabilities added incrementally, and that in the case of the F-35, the difference isn’t in trying to produce and field aircraft too soon, it is the presumption that we should be able to delay production until you get it ‘just right’ before you produce in any quantity. Unless you want to damage or kill a program, this has been shown to be counterproductive. If you had followed one of the links to past writings above, you would have found a reference to the report by a team of analysts from Center for Naval Analysis in the Defense in ATL magazine (link):

Our results (located at [link fixed by me], based on examining 28 programs across all Services, are very similar to those of the Congressional Budget Office and RAND [example] studies with one surprising exception: While from a purely statistical point of view we found that the relationship between both planned and actual concurrency and cost growth was very weak, in both cases, there seems to be a “sweet spot” of about 30 percent concurrency. That is, programs that plan on spending 30 percent of RDT&E funds while concurrently spending procurement funds actually experience the lowest average cost growth. Similarly, those programs that actually do spend about 30 percent of RDT&E funds while concurrently spending procurement dollars, even when not originally planned, also experience lower cost growth. Furthermore, programs with planned or actual levels of concurrency below 30 percent experienced higher cost growth than those with higher levels of concurrency. In other words, lower levels of planned or actual concurrency were actually worse than higher levels of concurrency. This is the complete opposite of what many in the acquisition community believe.

There is one sentence at the end of the paragraph I did not include above this time because I wanted to emphasize it:
We speculate that lower levels of concurrency may expose the program to higher levels of external changes.
Ya’ think? 

 Ground Rules and Assumptions

Some of these GR&As will apply to this post, but most will be saved for perhaps later discussions. It just so ‘happens’that I’ve been looking at the Post-WW2 aircraft program data and histories for quite some time and have been using what I’ve found to develop a database of production quantities, service lives, and variant definitions for various analyses. There is always variability in data definitions and data quality when comparing separate systems and programs over long time spans. There are also often gaps in information. For this exercise I chose to use only programs where the aircraft were actually fielded. Where there were gaps in the data, I tried to select the most conservative approach to estimate. For example, if a production contract was signed in either 1951 or 1950. I would select 1951 for cost purposes. For another example, if it wasn’t clear if a variant was fielded in 1963 or 1964, I would select the earlier year for service life estimates.) Since how well a fighter performs or how long a fighter remains effective is not directly related to how long before it is completely phased out, I chose to use the point in time a fighter begins to be withdrawn from ‘front line service’ as the standard for calculating ‘front line service life’. For early jets, this typically involved first transfers to ANG or AF Reserve units, with the exception of Interceptors, for which the ‘front line’ mission was transferred with the aircraft to the Guard and Reserve. Aircraft ‘variants’ are defined as having a model designation change for older jets, but the distinction is blurred with the introduction of later ‘Block’ type designators.


The intent will be to ignore subsequent block upgrades and mods (thought they are the norm since early WW2), unless they involve a model (A,B,C,D etc.) designation change. This approach is selected because ALL aircraft receive upgrades over their service lives, but model changes tend to flag major capability improvements with major changes to aircraft configurations. We won’t be dwelling on costs or service lives in this post but will focus on typical aircraft evolutions, from the perspective of time and numbers fielded, beginning the ‘first of type’ production units through when the first ‘definitive’ units were procured.

We’ll cover the period from 1944 to 1973 in two sections, the first one “Buy Now – Fly Later” we’ll list the Air Force aircraft in the first decade after WW2 for which major production decisions were made before the aircraft even flew. In the second section, “Baby Steps”, we will highlight how many early versions of the jets were built and often discarded instead of upgraded before the definitive versions were decided upon. This will highlight how though configurations were frozen before early variants were contracted for, they were only building blocks to get what was really needed.

Buy Now – Fly Later

Before getting into the particulars of the history of ‘concurrency’ and graphically illustrating how buying large numbers of early versions of aircraft before the first (or more) definitive variants is the normal course of things, I think it will be helpful to first how many early aircraft production contracts in our sample were put into place before a ‘production-standard’ (or sometimes even a prototype) aircraft first flight even occurred. For the very early aircraft, it could be seen that this was the result of wartime exigencies, but only those very early aircraft. All citations are from "Post-World War II Fighters: 1945-1973". In order [brackets mine]:


The AAF definitively endorsed the P-80 on 4 April (2 months ahead of the XP-80A's first flight) with a LC [letter contract] that introduced the first production contract. This contract, as approved in December, called for two lots of P-80s (500 in each). Delivery of the first 500 was to be completed by the end of 1945; …

On 7 January North American presented a bold design based on the successful P-51. This design promised range, reliability, and less pilot fatigue (the two pilots could spell one another). The AAF endorsed it at once. In fact, a February letter contract to construct and test three experimental P-82s gave way in the same month to an order for 500 productions…


[In January] The AAF order covered 100 service test and production P-84 [ later redesignated F-84] airplanes-25 of the former and 75 of the latter. This was subsequently decreased to 15 service test articles, which were redesignated YP-84As. The production articles were correspondingly increased from 75 to 85 and redesignated P-84Bs. [The P-84 ‘mockup’ was viewed by the AAF for the first time the next month]


[20 December] Although the prototypes were still under construction, a production order was released. Unit cost of the first 33 P-86s [ later redesignated F-86] authorized for procurement was set at $438,999.00—more than twice the aircraft's eventual price.


Funds released by President Harry S. Truman in January 1949 enabled the Air Force to execute, during May of that year, a cost plus-a-fixed-fee contract amounting to some $48 million, excluding a fixed-fee of almost $3 million. The estimated costs stipulated in the contract covered modification of the second XF-89 (YF-89) and fabrication of the first 48 production aircraft (F-89As). [Note: the first XF-89 had severe development problems, flew little and was lost shortly after delivery of the second prototype]…


[October 1951] The Air Force Council pressed for the development of revised Sabre 45 [F-100]. This decision ran counter to the belief of key development personnel that the aircraft would not meet the simplicity and cost requirements, basic to a day fighter. To obtain quickly a new fighter that would substantially surpass the F-86, the Air Force Council also agreed with the Aircraft and Weapons Board's recommendations to buy it in quantity prior to flight-testing, even though this ran the risk of extensive modifications in the future…

Initial Contract Date 3 January 1952 The Air Force issued a letter contract for two F-100A prototypes…

First Contract for Production 11 February 1952 The Air Force rushed through a second letter contract to procure 23 F-100As with fiscal year 1952 funds...

Second Production Contract August 1952 Having found the revised mockup basically satisfactory, the Air Force directed procurement of 250 additional F-100As. 1953: The LCs, previously awarded to Convair, were superseded by a definitive contract. This contract, still based on the Cook-Craigie production plan, did not affect the number of aircraft initially ordered. Out of the 42 aircraft under procurement, several were earmarked for testing and two (F-102A prototypes) were scheduled for flight in October and December 1953, respectively...


The F-101 and F-102 which employed the Cook-Craigie approach (no prototypes) in the pursuit of trying to mature technology before committing to LARGE production quantities, while still committing to volume production as soon as possible. Subsequent jets of the original type were purchased in volume, in evolved forms as a result of lessons learned in operation and test.

Baby Steps

Some adjudication and ‘calls’ in the analysis had to be made, because the real world isn't tidy. For example, I elected to use the F-104G as the definitive model type, though the US never bought it, it was the most numerous and built upon all the prior developments. I didn’t include a lot of F-86 variants prior to the D model because they were really parallel efforts. The F-84F was different enough from the previous versions that if it had been designated during the F-106 era, it would have certainly been given a different number designation, but it was still the final evolution of a long line of F-84s.

There were quite a few other types of aircraft, but not bought in 'major' quantities (except for perhaps the F-86H and precursors but I didn't want to over emphasize the F-86). The most important thing to take from this chart is NOT that in the past, we built aircraft as best we could, learned from them, and made them better in the next iteration.

The takeaway IS that we fielded needed technology as fast as possible knowing we’d learn something new, or possibly fall short (without fear), or learn we needed different or just ‘better’ technology. We then incorporated those lessons learned to get the systems we needed. Most of the time those precursor aircraft had limited front-line service lives and were seconded or scrapped less than a decade after they were built.

Compare that approach with today’s approach; the one used for the F-35. A limited number of aircraft have been produced, with the intention of making them all (or nearly all) meet the baseline standard (Block 3) through subsequent modification. There will probably be around 200 aircraft (or fewer) produced before the first Block 3 plane is rolled out, far less than 10% of the currently planned total production run, and all but the most early of those jets will be upgraded to baseline standard via mostly software/component updates.

Even if the production ramp up hadn’t been delayed by playing the faux ‘concurrency’ card, there still would have been far fewer F-35s needing upgrade than obsolete precursor aircraft produced in fielding previous ‘major’ types. Stretching the program added more costs and more total risks, just fewer technical ones.
I can't emphasize enough that how we frame the concurrency question defines the concept and discussion in the public square. We must recognize that the detractors are playing games with the definition of concurrency to make the F-35 seem worse than it is and worse than predecessor aircraft programs.

This is easily demonstrated by looking at the F-16's evolution.  

I’ve noted multiple times around the web, with no credible rebuttal to date I might add, that there were 291 F-16 Block 1 and 5 deliveries before the first 'nominally' useful Block 10 was built. To keep perspective, the YF-16's first flight (official) was Feb 74, and the first definitive and fully capable Block 30/32 F-16s for the US first flew Feb 87. Counting all partner nation deliveries, approximately 1800 F-16s were delivered before the fully capable Block 30/32s. Until the Block 30/32, all the capabilities of the F-16 were less than what was envisioned by the planners (just not the so-called 'Reformers'). The Block 30/32s were the first F-16s with full Beyond Visual Range-engagement and night/precision ground/maritime attack capabilities. They were the first with full AIM-7/AMRAAM/AGM-65D/HARM capabilities. They were also the first with Seek Talk secure voice communications. Until Block 30/32, the F-16 was mostly a hot rod for knife fighting on blue-sky days. At Block 30/32 and beyond, it was what the users wanted in the first place. An ‘all-weather combat aircraft’ to the users, or what the so-called ‘reformers’ refer to as ‘ruined’. Fielding 1800 F-16s aircraft before you reach a 'baseline' in Block 30/32? Thirteen years after first flight? I’ve said it before and I’ll say it again: THAT is 'concurrent development'.
To varying degrees, the same phenomenon can be shown for the F-15, and the F-18's, just look a the program history and the rationales behind the differences in variants.

P.S. Sorry I couldn't get this post up before it aired in most places. 'Life' intervened. 

No comments: