Bug with planes heading north

[Charky]

Unfortunately, it's not that simple. Heading could incorrectly be reported as 0, even though the plane has speed and altitude.

Example:
http://www.flightradar24.com/2012-05-02/15:27/DLH55M

Thank you Jakob for this example.
never seen a plane going south and track zero with a usual air speed.
Do you know if the BDS registers are cleared from time to time ?

And sorry to all. I was looking for a solution.

ps.
there might be a solution. EUROCAE ED-102A around page 98.
if you need it let me know
2.2.3.2.6.3.6 “Heading Status Bit” Subfield in Airborne Velocity Messages - Subtype=3
The ―Heading Status Bit ‖ subfield is a 1-bit (―ME‖ bit 14, Message bit 46) field that
shall be used to indicate the availability of ―Heading‖ information as shown in Table
2-34.
Table 2-34: “Heading Status Bit” Encoding
Coding Meaning
0 Heading Data is NOT Available
1 Heading Data is Available
2.2.3.2.6.3.7 “Heading” Subfield in Airborne Velocity Messages - Subtype=3
The ―Heading‖ subfield is a 10-bit (―ME‖ bits 15 – 24, Message bits 47 – 56) field
that shall be used to report the ―Heading‖ (in degrees) of the ADS-B Transmitting
Subsystem.
Range, Resolution, and No Data encoding of the ―Heading‖ subfield shall be as
shown in Table 2-35.
99
 EUROCAE, 2009
Table 2-35: “Heading” Encoding
Coding Meaning
(Binary) (Decimal) (Heading in degrees)
00 0000 0000 0 Heading is ZERO
00 0000 0001 1 Heading = 0.3515625 degrees
00 0000 0010 2 Heading = 0.703125 degrees
00 0000 0011 3 Heading = 1.0546875 degrees
*** *** ***
01 1111 1111 511 Heading = 179.6484375 degrees
10 0000 0000 512 Heading = 180.0 degrees
10 0000 0001 513 Heading = 180.3515625 degrees
10 0000 0010 514 Heading = 180.703125 degrees
*** *** ***
11 1111 1110 1022 Heading = 359.296875 degrees
11 1111 1111 1023 Heading = 359.6484375 degrees
Notes:
1. The encoding shown in the table represents an angular weighted binary encoding
in degrees clockwise from True or Magnetic North. The MSB represents a bit
weighting of 180 degrees, while the LSB represents a bit weighting of 360/1024
degrees.
2. Raw data used to establish the Heading Subfield will normally have more
resolution (i.e., more bits) than that required by the Heading Subfield. When
converting such data to the Heading Subfield, the accuracy of the data must be
maintained such that it is not worse than ±½ LSB where the LSB is that of the
Heading subfield.
3. The reference direction for Heading (whether True North or Magnetic North) is
indicated in the Horizontal Reference Direction (HRD) field of the Aircraft
Operational Status Message (§2.2.3.2.7.2.13).

[Mike]

As ADS-B is not mandatory there are quite a lot of errors, and no one seems to care. As I have never been in contact with a transponder I cannot explain the errors. But during the 5 years we have been running FR24 I have learned some patterns.
Many B777-200 have speed and heading 0.
Many RJ70/100 are landing some kilometers off the airport.
Many Turbo prop have speed and heading 0.
Some B747-400 have callsign problems.
One Antonov 148 is often flying backwards.
One SAS A340 is flying in Mach 2. (has been flying so for 2-3 years, why don't SAS fix this????)
One Aerosvit is flying with different strange transponder problems.
There are lot's of aircraft flying with incorrect/wrong/old HEX-code.
3 aircraft on Maldives are flying with the same HEX code.
Sky Airlines have more position transmission problems than any other airline.
About 10-15 aircraft are blocked from showing up on FR24 as they always transmit incorrect position data.

[Frank B]

Unfortunately, it's not that simple. Heading could incorrectly be reported as 0, even though the plane has speed and altitude.

Example:
http://www.flightradar24.com/2012-05-02/15:27/DLH55M

Okay. Perhaps I am misunderstanding how this works. I was under the impression that either the transponder is operational, and transmits all data, or it doesn't transmit anything at all. Is there an independent transmitter for each type of data?

[jakob]

What I would do is test for speed and altitude both being zero, and then assume that a heading of zero is a fault. In all other cases I would assume that zero heading means north.

Unfortunately, it's not that simple. Heading could incorrectly be reported as 0, even though the plane has speed and altitude.

Example:

http://www.flightradar24.com/2012-05-02/15:27/DLH55M

[Mike]

It's not safe to assume anything and there is not so much to test, you will have aircraft pointing in northern direction and flying in another direction. It must be better to show no direction, than wrong direction?

[Frank B]

What I don't understand is this: If you receive correct data for altitude and speed (non-zero values), why do you then assume that the heading is wrong just because it is zero? I could understand it if you didn't receive any data at all, but with everything else working, isn't it safe to assume that the heading is working too?

What I would do is test for speed and altitude both being zero, and then assume that a heading of zero is a fault. In all other cases I would assume that zero heading means north.

[Mike]

We haven't changed anything in this code from last version of the map and as I wrote, it's not so easy as Charky writes.
There are aircraft flying on FL100 with correct speed, and 0 heading. There is no rule or easy way of knowing if 0 is error or north apart from estimating heading from previous positions. Maybe that is something we can look on in the future.

[Frank B]

I agree with Charky. It doesn't seem to be that much overhead to do a few simple tests for zero values.

It may be that most aircrafts use 360, but it is still a bug for those that don't. And I do actually think it is an issue, because it makes the whole project look substandard, when some data are not displayed correctly.

[Charky]

thank you for your answer Mike.
you dont need to make an analysis of every uploaded message.
You only need to put the correct condition into the code.
I try to explain it again.
track zero is a valid track. so shouldnt show up as error "?".
if speed = zero then plane is at the ground or the message which contains track and speed has not been received.
if a plane is flying at fl100 and track and speed = zero then its an error "?".
if a plane is flying at fl100 and track is zero and speed not = zero then the track is correct zero.
if a plane is flying at fl100 and track not zero and speed is zero then its an error "?"

In the old FR24 map it had been corrected for some month's but with the new map layout it was comming back.
have fun.

[Mike]

We have about 100 feeders upload thousands rows of data every second. It's not possible for us to make a real time analyse of the data coming in. As 0 is mostly used by aircraft with transponder problems we show (?). There are also aircraft with direction 0 and speed =! 0. As I wrote, most aircraft uses 360 when they fly in northern direction so this is not a big issue.

[Charky]

you are right. Software/programming error

http://forum.flightradar24.com/threads/2550-Planes-with-%28question-Mark%29

solution: if track and speed = zero then "?"

[Frank B]

Actually it does populate all data, it is just the software that interprets it wrongly. Zero doesn't mean unknown, it means north. I could understand it if other parameters went to zero, like the speed or altitude for instance, but when there is this ambiguity with the heading, then it should not be used alone to indicate a fault. It should be interpreted as north when all other data are present.

[Oblivian]

Ah yes, its as I suspected. When track = 0' Then it is automatically considered 'unknown' (or ? as on map) as it relies on all the information to be populated. Which is why when it reads 1 it appears as known again

[Frank B]

Here is one now: http://www.flightradar24.com/DLH1YL

This one even keeps the correct speed when it goes to heading zero.

[Frank B]

Got an example flight for this phenomenon?

This is an example of a partial squitter transponder.

http://www.flightradar24.com/2012-05-03/09:54/PST43

notice the Speed and Track are 0 (unknown) but the location is not.

Yes, but that one is not flying north, so it is probably correctly marked.

I will keep an eye out for a good example.