Originally Posted by
Trumpet_trousers
Wow! I think you should run it up the flagpole and see who salutes it, then get all your ducks in a row going forward! Any chance that you could explain that in simple English?
Apologies - I thought I was talking to grown-ups.
ILS (the management process) and LSA (the technical/engineering analyses) are a set of techniques used to understand the in-service consequences of design decisions, and to ensure that the interests of the future operating organisations are given due consideration during the design phase.
They also include techniques to guide the design of the "optimum" support infrastructure and (more usefully) understand what the cost/operation/support trade-off would be if an alternative support infrastructure design was used for pragmatic or external reasons (eg "if the optimum support infrastructure has no 2nd line what would the extra cost be if I wanted to keep my second line to provide opportunities for rest tours for rotated personnel which would keep their skills current"). The output of the process is all the data for repairs, spares, support equipment, tools, training, facilities etc etc
The ILS/LSA process is a list of optional tasks and studies, each of which should only be done where they can produce information that is actually needed. In fact the very first element is to identify which tasks are needed (and why) using "least is best" and "if in doubt leave it out" as the primary guidance. This selection is supposed to be done collaboratively by purchaser and supplier, but the final decision is made my the purchaser - in UK defence procurement that's either a serving officer or (less often) a civil servant, and frankly their effectiveness it patchy.
In the days when I was involved in this sort of thing the procurement was either a DLO-led one or a DPA-led one. In the DLO-led procurements the ILS/LSA tailoring would usually be pretty effective - the desk officer would have responsibility for actually delivering the thing in a working form and so had a good idea what was actually needed.
In the DPA-led ones it was usually less effective because the DPA IPTs only took a programme to the end of the design phase before handing it over to the DPA to actually deliver. The DPA desk officers would generally tailor-out almost nothing, for fear of being found to have missed something. Those programmes were expensive, and developed reams of data that were just filed away because no one actually needed them.
Anyway, the foundation of all the analyses is reliability data, and reliability data is statistical*, based on the probability of events occurring in fleets over large numbers of missions/years or whatever. If you try to apply it to a single item and/or a short timescale the results will have very little meaning.
Is that better?
PDR
* since you don't like "stochastic"