Hello.
Something odd is going on with Flightaware tracking my aircraft. It’s no longer being tracked. I’ve noticed the issue starting on October 24. FA only published 1 of 3 flights that day.
At first, I assumed it was an aircraft issue. Today I flew about an hour receiving traffic and no transponder/ADS-B errors reported. I’m in a well covered area by FA. I even flew over my FA antenna. I got an ADS-B performance report from the FAA. They picked me up just fine on today’s flight. Flight Radar 24 has me too. So, this leaves something not right with N456TS/A58A20 getting “published” on FA’s web site. What’s going on?
I see the raw data so it’s getting that far but yeah, the downstream stuff is not creating an adhoc flight from it.
I wonder if it is because it’s not convinced you are in the air (2600ft pressure altitude and 100kts is marginal, that could be a just-landed jet at a 2600ft field)
One recent-ish change is that piaware is much more conservative about air/ground indications from the aircraft; it now needs to see a message that is definitively “in the air” or “on the ground” before forwarding that info on, not one of the more ambiguous messages that say “either it’s in the air or I can’t tell”. In the raw data I don’t see that data coming through. Does N456TS have a weight-on-wheels sensor?
The Air to ground setting is controlled by IAS. I have it set somewhere around 40 knots. It was in air mode during the whole flight. 100 knots at 2,600 feet 29.92inhg pressure altitude is typical for GA traffic.
I should add that the FAA has performance requirements for the Air/ground data. Mine passed that test today according to the FAA’s ADS-B compliance report.
Ok, it might be good to catch the original messages and see what’s going on.
piaware/faup1090 specifically needs to see a DF11 or DF17 message with CA=5 before it considers the aircraft airborne (reliably).
CA=4 is ‘Signifies Level 2 or above transponder, and the ability to set “CA” code 7, and on the ground’
CA=5 is ‘Signifies Level 2 or above transponder, and the ability to set “CA” code 7, and airborne’
CA=6 is ‘Signifies Level 2 or above transponder, and the ability to set “CA” code 7, and either on the ground or airborne’.
and the accompanying text says:
When the conditions for “CA” Code 7 are not satisfied, Level 2 or above transponders in installations that do not have automatic means to set on-the-ground condition shall use “CA” Code 6. Aircraft with automatic on-the-ground determination shall use “CA” Code 4 when on the ground, and “CA” Code 5 when airborne.
It may be that without a physical WOW sensor you’re sending CA=6.
There are separate requirements for transmission rate and switching between airborne and surface positions that can use airspeed / radio altimeter as inputs, but AFAIK they don’t affect the selection of the CA code and the results don’t help with reliably determining if the aircraft is airborne. (They do help with deciding if it’s on the ground)
If you want to capture the original messages next time you’re out, you can do it by leaving something like this running:
$ view1090-fa --show-only A58A20 >log
Ok.
I’m assuming the software for the transponder is written right, as there are no FAA report issues related to this. There could be a problem there, however, I think it’s fair to put that possibility near the bottom of the list. It sounds like you’re using different rules then the FAA? Right now, I have a compliant system once I upgrade the GPS receiver. (I’ll replace it eventually; current is SIL/SDA=1). The rest of the system is fully compliant. The Air/Gnd system is set per FAA rules. This will be the norm as virtually no light GA aircraft have weight sensors, (nor will they post 2020).
If the CA code is transmitted different then what faup1090 deems correct, then the problem lies with faup1090, not the aircraft. As I’m assuming they built the avionics correctly to the spec.
I can try to record the data at the next flight.
But what confuses me further is, on October 24, my antenna and another user’s antenna both picked up 1 of 3 flights. Specifically, the final flight of the day. Why did FA accept only one of the three flights? Once again, Flight radar 24 picked it all up.
I did update to piaware 3.1 at some point recently, but I can say for sure, I didn’t do it in between flights that day. (I assume the other local users also didn’t simultaneously update on the same day).
Can you contrast any differences on that day?
Faup1090 is conservative because, empirically, bad data was causing spurious arrivals and departures.
It makes sense to improve the rules downstream so that flights like your GA flights are handled better when reliable data is not available, but going back to trusting unreliable data would be a regression.
Flightaware does not use the same rules as an ATC environment because they have quite different requirements.
I’d like to see the raw messages so I can check they are being interpreted as expected.
I raised an issue internally with the flight tracking team to work out why adhocs aren’t being created from the position data.
Is view1090 installed by default with dump1090-mut 1.15 (pre-July)? I’m having trouble finding it.
view1090-mutability probably. But I don’t know if a version that old would have --show-only
Found it and got it ready. Didn’t add the -mutability.
Flew today. I’ve got a 5.6MB text file available. Can’t post it all here - 218,102 lines. The flight showed up on FA today. Change on FA’s side? It’s showing “Capability: 6 (Level 2+)”, which I think is what you’re referring to.
Snippet:
*5ea58a2019bbb2;
CRC: 000000
RSSI: -3.7 dBFS
Time: 220601513788.00us (phase: 0)
DF 11: All Call Reply.
Capability : 6 (Level 2+)
ICAO Address: a58a20
IID : II-00
*2020021c7c23d4;
CRC: a58a20
RSSI: -3.8 dBFS
Time: 220601534656.17us (phase: 120)
DF 4: Surveillance, Altitude Reply.
Flight Status : Normal, Airborne
DR : 4
UM : 0
Altitude : 2500 feet
ICAO Address : a58a20
*2020021c7c23d4;
CRC: a58a20
RSSI: -3.9 dBFS
Time: 220601560853.25us (phase: 180)
DF 4: Surveillance, Altitude Reply.
Flight Status : Normal, Airborne
DR : 4
UM : 0
Altitude : 2500 feet
ICAO Address : a58a20
*a020021c0201000000000085aa72;
CRC: a58a20
RSSI: -3.7 dBFS
Time: 220601562110.33us (phase: 240)
DF 20: Comm-B, Altitude Reply.
Flight Status : Normal, Airborne
DR : 4
UM : 0
Altitude : 2500 feet
ICAO Address : a58a20
Comm-B MB : 02010000000000
*8ea58a2099085302400005e24a0c;
CRC: 000000
RSSI: -3.8 dBFS
Time: 220601603809.50us (phase: 0)
DF 17: ADS-B message.
Capability : 6 (Level 2+)
ICAO Address : a58a20
Extended Squitter Type: 19
Extended Squitter Sub : 1
Extended Squitter Name: Airborne Velocity
EW status : Valid
EW velocity : 82
NS status : Valid
NS velocity : 17
Vertical status : Unavailable
Vertical rate src : 0
Vertical rate : 0
HAE/Baro offset : 100 ft
*8ea58a206811c3cca2c7e7df7866;
CRC: 000000
RSSI: -3.9 dBFS
Time: 220601773893.92us (phase: 300)
DF 17: ADS-B message.
Capability : 6 (Level 2+)
ICAO Address : a58a20
Extended Squitter Type: 13
Extended Squitter Sub : 0
Extended Squitter Name: Airborne Position (Baro Altitude)
F flag : even
T flag : non-UTC
Altitude : 2500 feet barometric
Global CPR decoding used.
Latitude : 35.699020 (124497)
Longitude: -79.571743 (51175)
NUCp: 5
*8ea58a2099085402600005ae024f;
CRC: 000000
RSSI: -3.8 dBFS
Time: 220602133914.50us (phase: 0)
DF 17: ADS-B message.
Capability : 6 (Level 2+)
ICAO Address : a58a20
Extended Squitter Type: 19
Extended Squitter Sub : 1
Extended Squitter Name: Airborne Velocity
EW status : Valid
EW velocity : 83
NS status : Valid
NS velocity : 18
Vertical status : Unavailable
Vertical rate src : 0
Vertical rate : 0
HAE/Baro offset : 100 ft
You can mail it to me at oliver.jowett@flightaware.com, thanks! gzip it or something first maybe 
Capability 6 is indeed CA 6. So that’s the most likely trigger - the air/ground messages went away, both because of gradual upgrading to piaware 3 and because the downstream code stopped trusting air/ground messages from earlier versions. Then the current rules for deciding when to depart a flight aren’t sufficient to get all GA flights right without an air/ground indication.
e-mail sent. N456TS.tar.gz attached. I don’t get why sometimes it posts to the web site and sometimes it doesn’t.