Page 1 of 12 12311 ... LastLast
Results 1 to 10 of 122

Thread: Discussion about Flightradar24 database

Hybrid View

Previous Post Previous Post   Next Post Next Post
  1. #1
    Team FR24 Mike's Avatar
    Join Date
    Feb 2010
    Location
    Sweden
    Posts
    3,110

    Discussion about Flightradar24 database

    If you want to discuss posts or data in
    https://forum.flightradar24.com/thre...craft-database
    or
    https://forum.flightradar24.com/thre...craft-database
    feel free to discuss here instead of the threads above.

  2. #2
    Super Moderator
    Join Date
    May 2011
    Location
    T-NZCH1, PP:PH New Zealand
    Posts
    5,057
    I note a lot of users also 'guessing' ICAO airline codes based on their name

    Wrong wrong WRONG.

    There appears to be no official search page like the aircraft type verification - https://www.icao.int/publications/DO...es/Search.aspx

    The closest comprehensive lookup/list (also, user contributed so may contain errors) - http://www.avcodes.co.uk/airlcodesearch.asp

    This does not include localised administration approved callsigns as these may conflict with international approved/established codes

    I believe FR24 can use these if evidence provided
    For instance in New Zealand: https://www.caa.govt.nz/assets/legac...oved_Signs.pdf
    SDA = Sounds Air
    Full Name: Sounds Air Travel & Tourism Ltd.
    IATA Code: S8
    ICAO Code:
    ICAO Callsign:
    Posts not to be taken as official support representation - Just a helpful uploader who tinkers

  3. #3
    Captain
    Join Date
    Feb 2011
    Location
    SBCT
    Posts
    522
    When are you going to set the correct icons for the types? You have literally hundreds of different types in the database that are set to the UNID icon. 4 that I can see at the moment : T38, M600, HDJT and SF50. These are just 4 out of hundreds, maybe even thousands. You already have the icon shapes on file so how hard is it to copy and paste them across?

    The fat icon should be reserved solely for UNID frames so that those of us that like to try to tie them up can spot them easily. Trying to do this in the USA when 50% of all icons are fat icons makes this annoying.
    Last edited by rik130; 2018-11-13 at 16:11.

  4. #4
    Captain
    Join Date
    Nov 2011
    Location
    Vergiate Italy
    Posts
    3,730
    Following on from Rik130 comment which nobody answered.


    We can't just set or update an icon. Correct icon is set based on the ICAO Type code. When correct ICAO Type code has been provided, the icon will be correct.


    So from the above statement (first page of new additions) your ICAO type to icon matrix/data table (what ever you use) is out of date by year(s)

    Is there any intention to update it? You have managed to update some types e.g. A339, A337 and A338 for example, when there are only a handful of examples currently flying compared to HDJT etc...

    Notice I asked nicely no need for mods to delete

  5. #5
    Purser
    Join Date
    Mar 2014
    Location
    F-ENBO3, F-ENRS1
    Posts
    181
    I agree with the two previous posters. Is anybody at FR24 working on connecting ICAO type designators to icons? There are probably a few 100 that are missing.
    Maybe I could do it?

  6. #6
    Captain
    Join Date
    Nov 2011
    Location
    Vergiate Italy
    Posts
    3,730
    Can someone explain how registrations only are added to the DB?

    I have noticed that just registrations no other details are being added to the dB. These are not ones from registers, such as Germany or France where you can work out the reg from the hex or a mass import such as from the FAA.

    When the reg is added to the DB how do you know they are correct? Are they just being taken from the callsign which implies they are correct?

    Not a complaint just asking, its actually quite a good idea providing that correct details are picked up.

  7. #7
    Team FR24 Mike's Avatar
    Join Date
    Feb 2010
    Location
    Sweden
    Posts
    3,110
    Quote Originally Posted by apgphoto View Post
    Can someone explain how registrations only are added to the DB?

    I have noticed that just registrations no other details are being added to the dB. These are not ones from registers, such as Germany or France where you can work out the reg from the hex or a mass import such as from the FAA.

    When the reg is added to the DB how do you know they are correct? Are they just being taken from the callsign which implies they are correct?

    Not a complaint just asking, its actually quite a good idea providing that correct details are picked up.
    Currently only US data is imported automatically but we are working on adding more countires for automatic updates. So outside of US all data is added manually today.

  8. #8
    Captain
    Join Date
    Nov 2011
    Location
    Vergiate Italy
    Posts
    3,730
    Quote Originally Posted by Mike View Post
    Currently only US data is imported automatically but we are working on adding more countires for automatic updates. So outside of US all data is added manually today.
    Thanks Mike

    If they are manually added then surely its better to add all the details? If the details cant be proved then how do we know the registrations that are being added are correct? I can see errors, registrations have been given to ground vehicles and other real aircraft have incorrect registrations applied (without any other information)

    A good example here, one of your guys has recently given a registration to a ground test signal

    Capture.JPG

    As I said if your team are adding correct registration data, that's good and helpful. But they need to check what they are adding is correct data to start with.

    Can we add registration only data if we know it to be correct and the type not known?

  9. #9
    Team FR24 Mike's Avatar
    Join Date
    Feb 2010
    Location
    Sweden
    Posts
    3,110
    Quote Originally Posted by apgphoto View Post
    Thanks Mike

    If they are manually added then surely its better to add all the details? If the details cant be proved then how do we know the registrations that are being added are correct? I can see errors, registrations have been given to ground vehicles and other real aircraft have incorrect registrations applied (without any other information)

    A good example here, one of your guys has recently given a registration to a ground test signal

    Capture.JPG

    As I said if your team are adding correct registration data, that's good and helpful. But they need to check what they are adding is correct data to start with.

    Can we add registration only data if we know it to be correct and the type not known?
    Of course it's best to have all data, 100% correct, real time, no errors and so one. But let's be realistic. We currently have 1.115.005 rows (1.1 million) of data (aircraft) from 250 countries. Less than 10 of these countries have official databases containing the ModeS/Hex-code. Probably 200+ countries don't even have an public official aircraft registration database. Probably 100+ aircraft are built, scraped or change ownership every day. According to FAA up to 30% of US registered aircraft are transmissing incorrect ADS-B data. Today we have 16.511 ModeS/Hex-codes tagged as incorrect. No one is able to check, verify or confirm all data that is available, submitted or processed. The database is maintained by 20+ people doing their best. If you find incorrect data please report it.

  10. #10
    Captain
    Join Date
    Feb 2011
    Location
    SBCT
    Posts
    522
    Quote Originally Posted by Mike View Post
    Of course it's best to have all data, 100% correct, real time, no errors and so one. But let's be realistic. We currently have 1.115.005 rows (1.1 million) of data (aircraft) from 250 countries. Less than 10 of these countries have official databases containing the ModeS/Hex-code. Probably 200+ countries don't even have an public official aircraft registration database. Probably 100+ aircraft are built, scraped or change ownership every day. According to FAA up to 30% of US registered aircraft are transmissing incorrect ADS-B data. Today we have 16.511 ModeS/Hex-codes tagged as incorrect. No one is able to check, verify or confirm all data that is available, submitted or processed. The database is maintained by 20+ people doing their best. If you find incorrect data please report it.
    So who are these mysterious other 19 database editors to which you keep referring in these discussions? I've been here for umpteen years and the only editor that's ever shown their presence on the forums is Maxi. Given the complexity of the database and tendency for errors/typos, why have none of these 'other' 19 editors pariticipated in the forums to clarify the accuracy of submissions etc? These questions were raised a few years ago when Maxi went on holiday for a month and absolutely no additions or corrections that had been submitted during that period were added to the database until Maxi returned and did them himself. You claimed to have 20 database editors so what do these other 19 editors do as it's clear they don't do any hex code additions or corrections?

    It's all well and good bragging about how many hex codes you have in your database but it's moot if you don't have an experienced team of editors screening the data to confirm its accuracy before sending it to the database. By your own admission you have 16,511 erroneous entries - and those are just the ones you know about. Those would not be in the database had they been screened beforehand.

    I can quite well imagine you view apg and myself as being a perpetual thorn in your side for giving you brief over the database but it's simply born out of sheer frustration when submitted corrections are ignored or overruled by your 20-strong editorial team who apparently know better, or additions are added but then overwritten back to the original incorrect data by your third party data sources and the whole cycle just keeps repeating. I've seen this happen numerous times with my Boeing 737 additions and every time I have raised the issue it's been completely ignored so I no longer bother posting them. What's the point in me wasting my time?

    Without wishing to create a cacophony of trumpet blowing you have the best hex codes researchers in the world right here in front you. I defy you to find anyone who has a higher attention to detail and level of screening/proofing the codes for accuracy than apg, 40612 and myself (+ a couple of others who don't post on these forums). You won't find anyone better because they don't exist. Everyone else simply works on the premise that (eg.) if the mode-s callsign field shows the registration then that's what the registration is and sends it off to the database without even bothering to proof it and this is causing the database to be filling up with rubbish. You have a 1.1m hex code database to maintain which is a mammoth task but your whole business model relies on serving accurate data to the public so why wouldn't you want the very best researchers in the 'industry' on your team to ensure that the database is maintained to the highest standard possible? If FR24 was my company I would be grabbing all 3 people with both hands.

    What am I missing?

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •