Time is always critical in PNT. Our GNSS Receivers make measurements on radio waves that travel at the speed of light. Any measurement errors feed through into position errors at a rate of 1 nanosecond (10-9 seconds) equals about a foot (30 cm). When high-rate Communications systems have timing demands at the micro-second level (10-6) this is 1000 times less demanding, but needs to be kept under control.
I have encountered a very different timing conundrum in Court, however: Time had been recorded by a GPS Device and was used to "prove" that a person was in a particular place at a particular time. The Device was accompanied by no specifications and no performance proofs. Nobody had tested that it delivered the famously accurate "GPS Time", and yet it was used to support a conviction. Since I have been deeply involved in GPS Receiver development, where "time" is notoriously tricky (those of you with an engineering background will know that unexpected singularities and negative time can cause mayhem in software!), I argued that time might have been "GPS Time" but was equally likely to be from an Onboard Clock, "steered" to GPS time periodically, with an accuracy that nobody in Court and nobody who had provided evidence to Court could possibly know! I hope that the time will come when we can all be a little more robust in our understanding of time and where it comes from.