The first two production F-35s (AF-1 and AF-2) arrived at Edwards AFB on 17 May 2010, i.e. over five years ago (
link). F-35s had clocked up over 15,000 flying hours by April last year (
link) so the figure is liable to be significantly higher by now. There have been no incidences of in-air catastrophic failure as far as I can tell and the level of system instrumentation and monitoring has been unprecedented.
Serious question: Wouldn't that provide sufficient data to start populating a MTBF database with some degree of confidence?
My I offer my insights?
1. AF-1 and AF-2 were test aircraft without all of the F-35s complex systems. Those systems are certain to drive the availability and logistics support tail of the aircraft.
2. Even if those two aircraft were equipped with all the F-35s systems, it takes much more than two aircraft to generate the kind of data required by a predictive PBL program. For starters, the test and evaluation environment is nothing like the operational environment. PBL must predict and control the entire supply chain, which involves much more than simply having reliable MTBF data.
3. Early MTBF data is notoriously unreliable. It takes years to mature a product and the processes that build and support that product. The C-17 struggled with that for years, and many of its systems were already mature (the engines, major electrical and hydraulic components, radios, comms, etc). I can't think of a single system/subsystem on the F-35 that is not brand new. And a lot use newly developed technology that has no past history.
4. A huge part of PBL is the diagnostic systems used to detect and isolate faults, the systems used to repair those faults, and the systems used to recertify the part to return it to service. These systems take years to develop and mature. Just "swapping boxes" when they fail will kill a PBL program dead.
5. PBL must also take into account the operational environment. The logistic system that enables the contractor to guarantee (for example) 80% availibility in the continental US is very different than the logistics system in continental Europe, which is different than the logistics system when deployed to Iraq or Afghanistan, which is different than the logistics system on board a carrier at sea, which is different than....you get the idea.
6. Tthe operational environment drives failure rates of many systems and components. For example, the salt air environment on a carrier will cause failures and generate servicing requirements that do not exist when operating out of New Mexico. The heat and sand environment in the middle east will drive different failures than those at sea. The cold in northern environments (Alaska, Northern Europe, etc) will result in different failures and failure rates. The stresses of carrier takeoffs and landings will drive all sorts of failures in the landing gear not experienced by USAF aircraft. And so it goes. It takes years of real world experience to generate the database that makes PBL possible. And each time you get a new customer or new operational environment, the database must be adjusted to take that into account. The database and the logistics system must also accommodate different block builds of the aircraft over time. So nothing is static and requires constant review and updates.
I hope this clarified.