Sure. So you take half the staff and put them on testing a second plane. Then each individual plane takes 10 years to test instead. (Remember that you don't gain person-hours for free just by shuffling people around.)
What would this accomplish? Instead of a total system time average of 10 years, you now get an average testing time of 10 years to which you still need to add the queuing time, so you're even worse off than before.
(The queueing time won't be five years with two servers in parallel, and I can't do the exact approximations in my head, but it'll be at least two years. In other words, by testing in parallel you worsen the cycle time from 20 years to at least 22 years.)
This is a good general rule: by taking on more work in parallel, you'll make the turnaround time worse. This is why lean consultants go on about limiting work-in-progress.
Also a call to learn some basic queuing theory! It comes in handy often.
When there's a defined process that appears to run in isolation, I don't see why there should be only one queue in this case, considering that the task length cannot be easily reduced.
I'm all for increasing the FAA budget so they can do their jobs better. I think that's a net positive for the industry.
What I was saying with my previous comments was that giving the current budget levels, it doesn't help to shuffle people around (without some strong assumptions on the process, which in my experience rarely are true in practise.)
Only if the workload doesn't shake out favorably in light of Amdahl's law.
Most mechanical processes aren't necessarily conducive to parallelization. Verification and information processing on the other hand can do favorably in the presence of non-reliance on a physical system-under-test.
Which subsystem vetting arguably is. If you're talking vetting specs.
Yes, there are some circumstances under which it makes sense to parallelise some of the work.
I'm assuming the FAA does this already. Human organisations have a bias to parallelise to a fault. Increasing the parallelism level beyond this does not improve lead times.
