No Land 3 says:
I think Trimotor came closest to a practical suggestion here. But I keep coming back to my point that any check has to be calculated for each take-off using the same variables as the original take-off calculation. IE it will vary for type variant, weight, departure point, slope, rwy surface conditions - the list is long!
So it will be subject to the same input error and be pointless.
If the input is correct then the performance will be correct. I suggest we are concentrating on the wrong end of the problem.
I agree that the time or distance to a selected speed has to be calculated for each takeoff, and if the correct numbers are put into the calculation all will be well. The point I was trying to get across in my original post was that some redundancy, an extra slice in the stack of Swiss cheese, can provide protection against wrong or incorrectly entered data.
If all the numbers provided are realistic and are put into the system accurately, as happens almost always, then the check is just that, verification that things are as planned. However when something goes wrong, when the passengers are all Sumo wrestlers rather than teenage gymnasts, or there is more slush on the runway than expected, or that supposedly empty container has been stuffed with a few tons of some illicit substance, or ..., then a check on time or distance to speed can save the day.
I suggested use of V1 as the check speed, simply because it is something already calculated and monitored. There is no reason in principle why some lower speed could not be used. Whatever speed might be used, the way to choose it should be decided by the owners of the SOP, taking into account input from people with appropriate ergonomics expertise and the PNF workload during takeoff.
I suggested basing the check on time rather than distance since the equipment required is miminal. Roll distance is not so easily measured, but could require extra hardware or modification of the software of existing systems. A conceptually simple system would be to adopt automotive technology and count revolutions of the landing gear wheels.
To drift slightly further from the main topic, one post a few pages back suggested that measuring time or distance to speed could give a reasonably accurate estimate of the weight of cargo and passengers, and averaged over a large number of flights be used to update the average wight of passengers and their carry-on baggage. Could the same information already be extracted from correlation of data captured by the flight data recorder and the loadsheets?