One of the joys of these blogs is that I can repeatedly show that life is more complicated than you would first think. This blog post follows on naturally from the previous one, in which I clearly stated that you can’t measure something before it happens.
No-one contacted me to protest at this bold statement, so I can assume that you think it is true. Today we will look at a case where it falls down.
Looking at the Data
The case in question comes from a helicopter and here is a part of the flight which shows the aircraft descending to land. The altitude and airspeed reduce progressively.
The altitude above airfield level is in grey and the airspeed in blue. You can tell it’s a helicopter because the airspeed reduces to zero before the altitude does. It’s the other way round for fixed wing aircraft.
Now, we can be fairly sure where the point of touchdown occurred – it’s about where the altitude reaches zero.
Gear On Ground
The aircraft also records the Gear On Ground signal to show when the wheels were in contact with terra firma. Let’s plot this on the same chart against (trust me about this) the same timebase.
Yes, the wheels touch the ground at 03:48:16 and the altitude reaches zero at 03:49:09, a full 53 seconds difference.
This effect is repeated at other takeoffs and landings, as illustrated below.
Here the gear on ground signal switches to Air at 03:01:19 and the takeoff actually occurs at 03:02:38. Even longer, at 79 seconds different in this case.
We have two questions to answer. Firstly, is the gear signal early, or are the rest of the parameters late? Secondly, what causes this discrepancy?
All of the information we have is recorded, and we have no independent measurement of the time that things happened, but as every other parameter operates in harmony with the others (e.g. when the aircraft rolls, the heading changes etc.) we can assume that it is the gear on ground signal that is in error. Also, and here I have the benefit of prior knowledge, we have seen this type of error before and we know the answer, so it really is the gear signal that is early.
How can the gear on ground signal operate about a minute before the aircraft actually gets there? The answer is that this data comes from a system where the data is recorded on the aircraft in the equipment manufacturer’s proprietary format. After the flight the data is extracted and converted from this special format into the HDF5 format before it reaches us. Somewhere in this conversion process some parameters can be misplaced in time by, usually, a minute.
The solution is that we work with the equipment manufacturer to resolve these issued before putting the finished system into service. Still, it just goes to show that you can’t take anything for granted in the flight data world.