Les inquiétudes d'“Aviation Week”

Qui ne connaît, chez les spécialistes du domaine, le sérieux et la compétence de Aviation Week & Space Technology (AW&ST), “bible“ des affaires aérospatiales, — selon l’expression consacrée? Lorsque AW&ST prend sa plus belle plume pour appuyer un article d’un éditorial et sermonner un ou plusieurs des grands industriels US de l’industrie de l’armement, c’est qu’il considère que l’affaire est sérieuse et qu’il a à la fois des arguments et des indications pour justifier son intervention. C’est donc le cas pour la question du JSF dont nous parlons dans notre F&C de ce jour.

L’éditorial concerne en fait Boeing et Lockheed Martin, pour leurs programmes 787 et F-35 (JSF). AW&ST reproche à ces deux compagnies de prendre des risques importants, voire considérables, en racourcissant le volume et la durée des essais en vol de leurs deux programmes. Ainsi, Boeing et LM risquent de ne pas rencontrer ni d’identifier avant un stade avancé du programme et en castrophe les “unknown unknowns”; il s’agit de ces difficutés inattendues et imprévues, dont on n’imagine même pas qu’elles puissent surgir, qui sont inhérentes à tout nouveaux grand programme de technologies avancées. AW&ST est encore plus inquiet pour LM, qui agit sous la contrainte des économies forcées et présente ainsi une situation où les pressions se combinent de plusieurs côtés à la fois pour resserrer jusqu’à les réduire à peu de choses les moyens et les capacités de contrôler le programme. Le danger est réel et il menace l’équilibre fondamental du programme.

«…in Boeing’s compression of the test and certification schedules for its 787, an aircraft whose novel design suggests a more complex and longer proving program may be prudent, rather than what Boeing is targeting, the shortest in history. Lockheed Martin also is proposing to compress the test program for its F-35 Joint Strike Fighter, but for another reason—instead of minimizing schedule slippage, the company is trying to deal with cost overruns.

»Regardless of motives, there is risk in each of these paths. Each company has much to lose and, in Boeing’s case at least, the prospective gains don’t seem commensurate. Both companies need to be watched closely by their customers: the airlines that will receive the first 787s and the 10 nations, mainly the U.S., that have signed up to buy at least 3,000 F-35s. At stake are not just huge investments and years of work, but also lives.

(…)

»Lockheed Martin, too, has a rationale for its proposed F-35 test schedule. It is trying to make up for a shortfall of about $600 million in program reserve funds, and the biggest cost-saver in its recovery plan is cutting two of the 21 ground- and flight-test aircraft from the program. The company intends to streamline flight tests, increasing the number of test points to be achieved in each sortie. And it believes that lessons it learned in its F-22 Raptor program, particularly the need for robust software testing and software’s potential to disrupt flight tests, will serve the F-35 well.

»The common factor in these improvisations is the same as Samuel Johnson’s definition of a second marriage — the triumph of hope over experience. Boeing and Lockheed Martin hope they have done their work right so far — no mistakes, no wrinkles, no glitches — despite experience that shows that often this isn’t so. The term “unknown unknowns,” restored to wide circulation in recent years by former U.S. Defense Secretary Donald Rumsfeld, goes back as far as aerospace itself. It refers to problems that are unfathomable as well as just plain unexpected. The unknown unknown is not something you thought you had gotten right but didn’t. Rather, it is something you didn’t even know you had to get right.

»What will be the 787’s unknown unknowns? The F-35’s? There’s no telling. They’re unknown.

(…)

»Boeing has less to lose here […] The F-35 program may be more problematic. The Air Force, the Defense Dept. and the nine other nations that have committed to the aircraft are all worried about how much it will cost, and they’re unlikely to forgive further cost overruns or big surprises in testing.

»When Air Force Gen. Sam Phillips led development of U.S. ICBMs in the 1960s, he used to say that a smooth, problem-free test program showed that you probably weren’t trying to accomplish enough. That’s debatable, but experience has shown that a test program — even in this age of computer-aided design — continues to serve a critical role in the development of new aircraft. It is no place to cut corners, tempting as it may be, for the sake of schedule.»

Mis en ligne le 8 octobre 2007 à 13H06