Could you explain to what degree incidents of unexpected interference result in dangerous failure modes? Is that sort of follow-up analysis done?
Usually dangerous interference comes through hard-wired connections: a device that feeds back frequencies into the power system, which then disrupts other devices; or a device that puts noise on an avionics bus, disrupting communications on that bus.
However, sometimes a device, even a device not designed to transmit, can create significant RF interference. For example, a device that has an internal power supply or transformer of some sort. If that power supply is not working right, it might generate an oscillating signal. Maybe even the device itself is tolerant of a certain amount of oscillation in its power and so is not affected by this. Then, there just happens to be a wire inside the device, connected to the power supply that's just the right length to act as an antenna at that frequency, and now you've got an RF transmitter on your aircraft that you didn't expect or want. There are bazzillions of wires in most medium-to-large modern aircraft, and some of them are bound to be the right length to act as an antenna and receive that signal, passing interference into who-knows-what.
That example may sound highly unlikely, but it does happen.
Yet, turning off consumer electronics is enforced by the honor system? Is there something else going on that makes consumer electronics unlikely sources of interference that would justify this lazy approach to safety? Why aren't all of these devices confiscated and locked in Faraday cages?
It just strikes me as odd that failure to pass an EMC SOFT grounds the plane until the issue is debugged, but then they just allow an random passenger to bring any random radio aboard just so long as they promise/remember to turn it off during critical phases of the flight.
That example may sound highly unlikely, but it does happen.
So highly unlikely that an honor system approach is justified?
As I've explained elsewhere in this topic, aviation mishaps are almost always the result of multiple factors, where if any one factor had been removed the mishap would not have occurred. "Passenger with unauthorized active RF transmitter" could be one source, but it's probably not going to cause a mishap alone.
In aviation we define "risk" as a combination of probability and severity. You can drive down the risk for a particular hazard by either lowering the probability or lowering the severity. The only way to completely eliminate risk is to not fly, so every flight involves accepting some level of risk. It's really a question of residual risk you're willing to accept balanced against the costs (financial and otherwise) incurred by attempts to further reduce the risk. Several people in this topic have discussed the costs of just the current system of having everyone turn off their phones during critical phases of flight, and complaining that the FAA doesn't apparently care about such things. I disagree with them--if the FAA truly didn't care, they would do exactly what you say: confiscate all electronics at the gate and lock them in a shielded box.
>It just strikes me as odd that failure to pass an EMC SOFT grounds the plane until the issue is debugged, but then they just allow an random passenger to bring any random radio aboard just so long as they promise/remember to turn it off during critical phases of the flight.
Keep in mind that I'm flying in a developmental test environment. That means that we're dealing with a lot of stuff that has never flown before, or are using it in ways it has never been used before. Because of the nature of what we're doing, we already have a lot of risk baked into our flights that an ordinary flight does not have, so we tend to be a lot more conservative about the risk factors that we can control.
[EDIT to add quote for context]