cogsim
As the design consideration is approached in real time, redundant systems simply fail together. If one system has sensors that have different designs, the chance of double or triple fail becomes remote.
Identical fuel systems in BA038 failed their TRENT 700's only seven seconds apart. The failure tracked identically as each engine spooled up, then starved, and settled, with very close EPR's.
And that is a complex system. A third engine, or two more, would the failure have been the same? Very likely. A redesigned Heat Exchanger, and the problem is solved. Pitots are extremely simple, yet they have remarkably vulnerable features. A narrow aperture surrounded with a sharp leading edge that ices faster than a blunt one.
If ICE is the consideration, three separate design approaches would likely produce three anomalous solutions to the goal. Redundancy works well if in a "fallback" position, so I like your anomalous heat solution. Exposing three identical devices to the exact same conditions will produce three failed devices. If three concurrent readings are desired, better each device has different characteristics.
This is why I never understood the need to "sample" and then ask for a vote.
Likely the results of the election will be the same. If not, they soon will be.
So this is why anomalous design works well when Redundancy is simply a siren song.
This is from memory, so apologies if it isn't what you saw. I think I inserted a snarky comment about "Backup systems". So tell Air France to cancel their Goodrich deliverables, purchase the BUSS from Airbus, and happy skies......
My apologies - I presumed that Bearfoil could see and reinstate the original post which he had deleted. FYI, mods can see a deleted post and reinstate it. All deletion does is hide it from general view. There is one caveat - delete post #1 and the thread disappears into the never never.In future, if you want a deleted post undeleted - simply ask and your wish will be granted. JT