An interesting post by Daniel Baker, CEO on the FlightAware forum. I suppose this was inevitable?
First, thank you for your support and contributions over the years............................
I'm writing today to discuss an issue with the FlightAware multilateration (MLAT) network. As many of you know, FlightAware not only operates a free MLAT network (currently using 136 top-of-the-line Xeon CPUs covering 113 geographic regions) for all PiAwares and FlightFeeders, but FlightAware also feeds back all the MLAT calculations that you've contributed to. This means that you can view the data locally on your dump1090 web interface or any other tools you have connected to your receiver. We're the only ones that do this and it's awesome.
Unfortunately, a very small percentage of users are re-distributing and aggregating FlightAware's MLAT calculations, often for the purpose of tracking aircraft that have specifically requested not to be tracked and are on one or more "block lists" used by the flight tracking industry.
As a result, we have come under pressure from aircraft operators with legitimate security concerns to stop feeding MLAT data back or remove their aircraft from the feed. Specifically, a lot of the concern originates from law enforcement, military, and other government operators ..............................
Read the full post and the proposed changes:
https://discussions.flightaware.com/post197041.html?sid=3091c8e23dec1db5db2473c64e9f0606#p197041
Quotefor the small number of aircraft on this list (less than 1% of flights, most of which are in the US, and none of which are airline flights), we'll continue to feed MLAT results to PiAware & FlightFeeder, but with an anonymized Mode S code and no ident. This means your maps will look the same and this has no impact on ADS-B flights or other Mode S data you're tracking locally, which are decoded and displayed entirely within your device, without the involvement of FlightAware's servers.
Can someone translate please?
Does this mean that the display on my PC will be the same, but the shared display on FlightAware web site will have anonymized data? And I must not share the date from my PC. If so, no problem ;D
Quote from: IanH on August 16, 2016, 08:26:31 PM
Can someone translate please?
As I understand it, the local Dump1090 display will include anonymised data for "blocked" aircraft as well as the "raw" data from the local Mode-S receiver. In theory we should still be able to identify the aircraft "locally" but not if the data is redistributed. But I've not yet seen this myself so could be interpreting it wrongly.
We'll have to wait and see how it's implemented. I think the goal is to stop FlightAware MLAT "blocked" data being published by other aggregators.
I've just updated my PiAware to 3.0.3, which has some new features that justify it.
- Aircraft type and registrations, when known
- Country of registration and flag
- Airframes hex link
- Total Aircraft count
- Some map customisation, including range rings
I've attached a screen comparison.
In my list of 172 aircraft, there was only one aircraft anonymised.
PiAware plotted the anonymised aircraft on the map but not the "associated" Mode-S record, which was hex ab9dc6, the CL60 N848CC of GAMA Charters Corp.
I think FlightAware's approach is preferable to the total "blocking" of ADS-B and MLAT plots that some tracking sites adopt. I can live with anonymisation, with an "s" or "z". ;)
[Attachment deleted by Admin to save file space]
Looks like FA have incorporated a lot of the features from the community setup script here: http://discussions.flightaware.com/ads-b-flight-tracking-f21/ads-b-receiver-project-setup-scripts-t36532.html
Output from my Pi (actually an "OrangePi PC") using previous version looks like this.
Screenshots in following order (any or all of the options can be toggled on or off:
- with tracks (and showing reg of selected)
with tracks and heatmap
max range recorded at various altitudes (can't remember what they are)
tracks and weather
[Attachment deleted by Admin to save file space]
The script also sets up a range of performance graphs as well as the Dump1090 map.
Usually fairly static but could be used to show changes using different antenna, gain settings, etc, The view below is for a month but you can show hourly, 6 hours, daily, weekly, monthly, yearly.
The script was initially for the RPi but there was enough interest and feedback (much from abcd567) to also make it work with the OrangePi.
Until the scripts work with version 3 of dump1090, I'm staying with this ;D
[Attachment deleted by Admin to save file space]
Thanks Ian.
There are useful features in the script which would be good to incorporate in the standard PiAware build. The majority shy away from making "manual" changes, me included.
For anyone prepared to install Raspian on an SD card (no different to installing Flightaware package) and accessing the terminal screen (either using a monitor or remotely via Putty), installing via the script is just a matter of typing in commands - actually copying and pasting is best. It installs dump1090-mutability and then some add on packages.
Not much different from the early days of installing dump1090 itself which was command line stuff ;)
Quote from: IanH on August 19, 2016, 12:13:08 PM
For anyone prepared to install Raspian on an SD card (no different to installing Flightaware package) and accessing the terminal screen (either using a monitor or remotely via Putty), installing via the script is just a matter of typing in commands - actually copying and pasting is best. It installs dump1090-mutability and then some add on packages.
That was enough to put me off. ;)
Installing PiAware is simply a matter of writing the downloaded image to the SD card. Easy peasy. No HDMI monitor (which I don't have) or USB keyboard required.
If you're manually updating Piaware via SSH (Putty) then the latest version of Piaware is 3.0.4
It might be important to you (or not) to note that installing Piaware3 over Piaware2 via the SD card resets your personal preferences, this includes resetting the password to the default one.
Seems a worthwhile update to already good software.
Quote from: Bethsalem on August 19, 2016, 06:25:43 PM
If you're manually updating Piaware via SSH (Putty) then the latest version of Piaware is 3.0.4
3.0.4 has only jut been released. Did a fast upgrade via my Stats page Command option - "Upgrade and restart PiAware"
Just realised that the tracks/range/heatmap are an additional addon to the Piaware setup script :-[ and not part of the package.
Installed a while back and I'd forgotten about that.
The performance graphs and registrations are part of the setup script.
This may be old news, but on Sunday FlightAware had an unspecified infrastructure anomaly which meant that PiAware feeders were not being sent mlat data back to them. This has been sorted out now, but there might be some feeders who still aren't getting any mlat data.
To sort out this issue you have to log into your account and go to settings and send the command to "Restart PiAware". This should get you back up and running with mlat fixes.
Thanks for posting this.
MLAT should be OK now. Mine are without needing to restart PiAware:
Originally: Multillateration results fed back via piaware are temporarily unavailable (some infrastructure issues that need sorting out).
1430Z: Normal service restored. You will need to restart piaware to start receiving results again; we will schedule automatic reconnects at a quiet time for those that haven't reconnected before then.
https://discussions.flightaware.com/post208675.html#p208675
Bethy
Thanks for that info. Looks like I was affected but hadn't realised it this morning. Wondered why occasional GA stuff going overhead wasn't showing. Just assumed no transponder.
Anyway Piaware restarted and lots more on screen now. Makes you realise how much MLAT contributes.
Ian
Thanks for this info. I was experiencing the same issue and as I only display Mlats on my ModeSMixer2 screen, it was obvious I was having a problem.
Followed Bethy's instructions and after a bit of stopping and starting of my ModeSMixer script it has sorted itself out.
Anmer - you can close out the FA support call now - thanks
Tim
Seems the auto reconnect hasn't been applied yet, so a PiAware restart is still required.
I've lost my return Mlats again this morning. Tried the restart PiAware command and stopped and started my MM2 script even rebooted the Pi. Nothing doing, no Malts :'(
Tim
Do you see MLAT on the Skyview map and list?
Not sure what the Skyview map and list are or where to locate them. If you mean via the FA web site, I never go there - until these last few days to restart the program.
If you can point me in the right direction I'll take a look.
Tim
Hi Tim
My MLAT data is missing and I see similar reports on the FA forum, so it looks like a server side issue. Houston is still asleep!
In your FA account, there's a top menu option "My ADS-B".
This takes you to your stats page from where you can view the local Skyview map and aircraft list. Depending on the Dump1090 install it might show as "View Live Data"
Anyone can view the stats pages and I think this is your one? Your own access will show more options.
http://flightaware.com/adsb/stats/user/tpq777#stats-19395
Hi Mike,
Thanks for that. I can see my data and it appears to show Mlats in the list, so I will sit tight until the guys with the keys open up and sort things out!
Tim
If you click the MLAT on the list (non-ADS-B aircraft), they're not on the map and the pop-up box doesn't show MLAT. Whatever plots the location doesn't seem to be working.
We'll have to wait for the Texans to get breakfast and head off into the office.
http://discussions.flightaware.com/ads-b-flight-tracking-f21/multilateration-results-temporarily-unavailable-2017-07-21-t39329.html
2017/07/21 0400Z: The problem has reoccured; results via piaware are temporarily unavailable again.
Still no MLAT this morning. Restarted Piaware with no difference. MLAT data being received by Flightaware according to stats, but MLATs in list show no location, so data not being fed back.
I see obj wrote yesterday:
"Yeah, this is a system-wide problem now; the band-aid from Monday did not stem the bleeding sufficiently. See the announcement post.
I am working on a proper solution but it is probably a few days away."
Although someone in Coventry reckons it was working this morning - what's your status?
Meanwhile back to PP for MLAT ;)
Quote from: IanH on July 22, 2017, 11:11:36 AM
Although someone in Coventry reckons it was working this morning - what's your status?
I checked earlier and no MLAT.
Checked again and I now have MLAT plots again.
I think a fix has been found and is being progressively rolled out.
Looks like it - MLAT are back now ;D
I'm told MLAT is working for everyone now.