PilotAware
British Forum => General Discussion => Topic started by: Easy Cruising on June 22, 2016, 03:08:42 pm
-
FLARM has been discussed quite a lot here, so I'm probably repeating the issue, but I still dont have a clear understanding:
Generally it seems to be that FLARM is some proprietary protocol and the developer won't share its decoding since that would presumably reduce sales of their expensive hardware devices. For me the existence of super-cheap ADSB send/receive modules like PAW suggests that FLARM's protocol will die, sooner or later. However, loads of gliders and light aircraft use it currently, so it would be good to pick those up. The frequency used by FLARM seems rather similar to ADSB (but I'm not techy in this area).
I keep seeing articles on internet which suggest that decoding FLARM is not so tough, and people are doing it :
http://comments.gmane.org/gmane.comp.mobile.osmocom.sdr/455
Is that really the case or there's some mega encryption on it or something that's indecipherable ? If not, whats to speak against a flarm decoder in paw ? It's not clear to me whether that is a technical or legal issue.
Anyone clarify our position (no pun intended) ?
-
The RF protocol is pretty standard, however the data within the packets is encrypted and there in lies the issue. Technically you can decrypt it, the system used isn't that strong but does change each year - you have a mandated software update yearly with FLARM that changes the keys / algorithm and if you don't do it you won't see anyone and they won't see you.
So it boils down to a legal issue, it is a proprietary and closed system so I would expect a fairly swift flurry of solicitors letters if the means to decode it was in anything other than a FLARM supplied bit of hardware.
It didn't used to be encrypted, but then people started to listen in and not by purchasing FLARM hardware, so they closed the door.
-
Just to follow up on Jeremys statement.
The means to decrypt FLARM is already in the Public Domain.
It would be irresponsible of PilotAware to build a product around a protocol which was not open, and could change at a moments notice.
This would upset our users if we said 'use this and see FLARM', only to have FLARM change the encryption, thus making them invisible once again to us.
the current strategy will be to provide a FLARM RS232 input to PilotAware.
This would allow you to purchase a FLARM Mouse at a mere £444
which could feed its data to PilotAware, and hence be provided to your Tablet display, along with
P3I, ADS-B, Mode-S (and potentially Mode-C)
Thx
Lee
-
Ok, got it. So the developer of FLARM is actively disguising its transmissions .. somewhat hypocritical from a company that's claiming to promote air safety, but that's business I guess.
Thanks for informing about the FLARM mouse. Still, I'd feel bad feeding money to them to help them prolong the life of their closed and expensive products.
-
Interestingly, I spoke to a glider pilot the other day, explaining PilotAware, how light and portable it is and was surprised when I found out he and many like him do not use Flarm at all as it is too expensive.
He was quite excited to hear about PAW Classic, and went away with web address details.
I wonder how many Flarm users there are out there, and have a suspicion that at those prices, many may defect to PAW anyway.
Clearly, those not using Flarm, due expense, are very excited and I wouldn't be surprised if we don't see the gliding community adopting PAW in some places.
Have you reached out to the British Gliding Association and British Parachuting Association with details of PAW?
-
Looking at the Flarm mouse I cannot tell whether it transmits also or only receives. Any idea?
-
I believe it transmits and receives FLARM.
It says it has an RS232 data out port so theoretically it can output data to a PAW when Lee writes the code to do it.
-
I suspect that most of the Flarm users are in Switzerland and Germany. You can get a good idea of the numbers from live.glidernet.org and how the numbers are distributed.
In the Uk at least there is a possibility that PAW could be more dominant.
-
I believe certain glider sites have a good uptake like Lasham where all the club gliders have it fitted. Certain military sites also have FLARM in their gliders and light aircraft. However, there is very little uptake among powered GA. This is where PAW is aimed, with a higher power transmitter than FLARM for longer range detection and to combat the shielding effects of a metal aeroplane and engine.
-
If my understanding of FLARM is correct, it is not a like-for-like comparison with PAW.
PAW detects other suitably electronically conspicuous aircraft and passes information about them to another device so that second device can display that information. So PAW is a traffic detection system.
FLARM seeks to be a collision avoidance system. FLARM seeks to detect and analyse the position and trajectory of other FLARM-equipped aircraft and compare that, using algorhythms suited to the manner in which gliders fly/are flown, with the position and trajectory of the user's own aircraft and determine if there is a conflict. Gliders often tend to fly in gaggles close together circling in thermals so the pilots of those gliders are often well aware of the presence of the other gliders but FLARM is providing that extra layer of protection against misjudging your own flight path relative to the others around you.
What I am trying to say is that FLARM and PAW are similar but were not actually designed to do exactly the same function.
And then there is PowerFLARM, which adds in detection of ADS-B/Mode C/Mode S, but that is a whole other story.
Steve
-
What frequency does FLARM Tx/Rx on? Is this receivable by either the RTL dongle or Bridge unit?
If this is the case then could PAW simply, without decoding, detect the presence of the signal and give some simple warning of gliders in the vicinity?
-
It's on 868.4MHz I believe, in the same band where the bridge operates.
However, it would require either the ADS-B dongle or the bridge to be retuned to check (if it's possible to retune the bridge on the fly, dunno) and then retuned again back to their original frequencies.
Alternatively an extra DVB-T dongle could be used, but probably only on the Pi 2 due to the far greater processing required to run two SDRs at the same time.
-
This says 868.2 and 868.4MHz.
http://comments.gmane.org/gmane.comp.mobile.osmocom.sdr/455 (http://comments.gmane.org/gmane.comp.mobile.osmocom.sdr/455)
There's some SDR stuff on FLARM here:
http://www.rtl-sdr.com/tag/flarm/ (http://www.rtl-sdr.com/tag/flarm/)
I think it's been mentioned elsewhere that the presence of a FLARM signal is detected by at least one other system, and this would be a bit like a Mode S non-ADS-B detection.
-
Has anybody asked FLARM to produce a suitable dongle - at an affordable price as the hardware is almost free?
-
You would need to go via a DVB-T dongle as the FLARM chipset is different to the Bridge, so the format of the RF packets are not the same.
Remember that the FLARM RF stage runs a very low power of 10mW with a stated effective range of 3km-5km. Using a DVB dongle for this will work, but with such low powers it doesn't take much interference/noise to swamp a signal. Within the FLARM they have filtering on the RF stage to help reduce this and use dedicated hardware akin to the Bridge to handle this.
Flying at a closing speed of 120kts, 3km gives you around 48s warning, at 5km this increases to 1m 21s, assuming detection is immediate but you could probably drop 10s-15s off those to ensure you have a series of received packets to do some processing with.
For Mode-S detection Lee compares known distances from other transmissions to assist with the bearing less targets, with FLARM what is there to compare to? Unless you just alert on every detection given the short range. Of course if you decode the packet you have the position of the FLARM transmitter and could take action as appropriate, but then sit back and wait for the lawyers to get involved.
Has anybody asked FLARM to produce a suitable dongle - at an affordable price as the hardware is almost free?
FLARM sell an OEM radio module for integration into other systems, pricing is on request under an NDA but I doubt it would be cheap (probably £100-£200 ish) plus the need to manage it's firmware updates too.
PS: Doesn't FLARM also use a frequency hopping RF system? Seem to remember reading that somewhere.
-
You would need to go via a DVB-T dongle as the FLARM chipset is different to the Bridge, so the format of the RF packets are not the same.
The proposal is that you'd just measure the signal strength, or just detect if a signal was present. Might detect garage door openers or something as well, but hey ho.
Remember that the FLARM RF stage runs a very low power of 10mW with a stated effective range of 3km-5km. Using a DVB dongle for this will work, but with such low powers it doesn't take much interference/noise to swamp a signal. Within the FLARM they have filtering on the RF stage to help reduce this and use dedicated hardware akin to the Bridge to handle this.
You can get an 868MHz filter here:
http://www.rtl-sdr.com/some-new-rf-filters-from-adam-9a4qv/ (http://www.rtl-sdr.com/some-new-rf-filters-from-adam-9a4qv/)
It wouldn't help with interference from the bridge transmitter though!
PS: Doesn't FLARM also use a frequency hopping RF system? Seem to remember reading that somewhere.
Is this the 868.2/868.4MHz thing?
-
Is this the 868.2/868.4MHz thing?
I don't think so, but it could be. I would need to look back at some notes on the FLARM RF front end I have knocking about somewhere. I've been re-organising my office/lab so know they are about, just got to figure out where they are!
Whilst I like electronic documentation, as it saves a huge amount of space, I have not yet scanned everything in...
-
It's been a while but I notice that the Stratux appears to be able to receive Flarm targets through its UAT receiver, if programmed to do so (although I'm not tech enough to know if it's really working air-to-air):
https://www.reddit.com/r/stratux/comments/6pa2nj/flarm_receiver_functionality/
Relying on special ground stations to relay traffic seems like a complex route to me, so this subject interested me. Might be something for PAW too.
Can anyone who understands these things confirm or debunk it ?
-
Flarms legal hounds would find it difficult to hunt the guys and individual enthusiasts on GITHUB who have implemented flarm on stratux.
However, Flarm do not allow use without first handing over several pounds of flesh for a licence (Flarm and a leg prices remember) and would almost certainly send Lee a letter along the lines of "stop or else" for unlicensed Flarm use.
I have powerflarm and PAW in my aircraft, the range is very poor flarm to my powerflarm (almost useless) and I prefer the significantly better range I get from OGN-R even thought the overage isn't 100%
Deker
-
Ok. Thanks Deker. OGN-R is non existent outside of the UK though. Flarm seems rather sinister somehow. Hopefully it'll die-out as ads-b solutions become cheaper and more widespread.
-
EC,
ADSB is not likely to provide the precise anti-collision alerts at extremely short range which gliders need to allow them to operate in close proximity (without alarms), e.g. while thermalling, yet get warnings of imminent collision when necessary - which FLARM is specifically designed to provide.
I have been involved with PilotAware since it’s early days, but I also run FLARM in parallel and integrated with PilotAware because there are two very active gliding sites in the area where I normally fly, which is outside the main area covered by the OGN-R network, though this area is expanding all the time and IMO OGN-R is definitely the way ahead - at least in the UK, though we do have a couple of fledgling sites in Mainland Europe.
Best Regards
Peter
-
Easy Cruising
Would you be interested in setting up an OGN-R
It is community driven and benefitted by all
We can help and subsidise
Thx
Lee
-
From what I understand about pilot aware and flarm, pilot aware tells you of similar equipped aircraft in your vicinity by transmitting your gps data to the other aircraft. Flarm on the other hand also transmits gps position and height of your aircraft but on receipt of a flarm signal from another flarm equipped aircraft they calculate using your data and the other aircraft's data if there is a possibility of a conflict and gives an warning accordingly to both pilots. This is very useful when multiple gliders are circling in the same thermal as it will only give an audible warning in the case of a conflict. Because of this feature I doubt that flarm will die out in the near or distant future.
-
I think exfire was trying to explain that also, but it seems to me that the information from a waas gps + pressure sensor is the only data available to any of these systems (flarm doesn't tap the air-data or attitude etc from the transmitting aircraft does it?). In this case it's only the software that decodes this data that differentiates the systems, so => build out the ads-b based software and flarm becomes redundant. SkyDemon already can do this well from my experience using it. I'm not tech enough to understand the details of the protocols.
-
EC,
FLARM is a bit cleverer than that. Each unit continuously calculates its own 3-dimensional track 20 seconds or so ahead (possibly further for PowerFlarm) and transmits this virtually continuously, automatically updating the data as it goes. It then compares its own track with any received from other FLARM units in the vicinity to calculate likelihood of collision. Having determined this, any likely danger is monitored and alerts issued as necessary. All this designed in such a way as to NOT issue alarms during ‘normal’ glider flight in close proximity unless the tracks indicate imminent collision.
ADSB is only designed to do this sort of thing at the sort of levels of installation normally found in CAT aircraft. I would think it unlikely that this will change significantly at least in the near future.
Regards
Peter