FLARM.....for good, or evil??

On 10/29/2010 12:51 AM, Andy wrote:
On Oct 28, 9:19 pm, wrote:

Also, to clarify, ADS-B does no path estimation of its own. That
function either would have to be added into an ADS-B unit by the OEM,
similar to the way Flarm does today - unlikely to be done in a glider-
specific way IMO - OR, it would have to be done by a separate external
device, perhaps a navigation computer/software like Oudie, WinPilot,
SN-10. For it to be effective manufacturers would all have to agree to
use the same algorithm, which also seems unlikely, unless they all
adopt the Flarm algorithm. That seems somewhat unlikely too, since I
don't think Flarm would want to start splintering how their algorithms
get used by splitting out the Flarm link technology from the collision
algorithm (which would have to be modified to accommodate the
differences in how path estimations get generated - with unpredictable
results). PLUS the external device OEM's would have to adapt to using
ADS-B inputs - another standards issue.

No matter how hard I try, it seems highly improbable that you will be
able to stitch together a satisfactory collision avoidance system for
gliders using ADS-B technology developed for general aviation. You'd
have to be satisfied with the simple functionality offered by ADS-B -
which would be fine if you generally come into conflict with GA and
airliners more often than other gliders, but there are a bunch of us
for whom the opposite is true. Then the problem becomes some gliders
using Flarm and others using ADS-B, you lose some of the Flarm
benefits of path estimation for the non-Flarm gliders.

9B

9B


You are probably correct that no one is going to beat FLARM in an
optimized collision avoidance solution for high density glider
environments. That's obviously their focus and they are good at it.

However, most recreational, non-contest pilots, primarily need a system
that will reliably alert them to other aircraft in their general
vicinity. If I enter a thermal and know that there are 3 other aircraft
in the area, and I only see two, I'm going to abort and go elsewhere. A
contest pilot obviously wants more data.

What is interesting about the Parowan situation is that this was not a
gaggle of gliders. It was two gliders who apparently did not have a
proper appreciation that they were near each other. A simple graphical
display that showed their relative positions, with a very simple
collision avoidance algorithm, or some form of auditory announcement
could have prevented this accident. That's not to say that the FLARM
simulation was not impressive.

--
Mike Schumann

On Oct 28, 11:20*pm, Mike Schumann
wrote:
On 10/29/2010 12:51 AM, Andy wrote:





On Oct 28, 9:19 pm, *wrote:

Also, to clarify, ADS-B does no path estimation of its own. That
function either would have to be added into an ADS-B unit by the OEM,
similar to the way Flarm does today - unlikely to be done in a glider-
specific way IMO - OR, it would have to be done by a separate external
device, perhaps a navigation computer/software like Oudie, WinPilot,
SN-10. For it to be effective manufacturers would all have to agree to
use the same algorithm, which also seems unlikely, unless they all
adopt the Flarm algorithm. That seems somewhat unlikely too, since I
don't think Flarm would want to start splintering how their algorithms
get used by splitting out the Flarm link technology from the collision
algorithm (which would have to be modified to accommodate the
differences in how path estimations get generated - with unpredictable
results). PLUS the external device OEM's would have to adapt to using
ADS-B inputs - another standards issue.

No matter how hard I try, it seems highly improbable that you will be
able to stitch together a satisfactory collision avoidance system for
gliders using ADS-B technology developed for general aviation. You'd
have to be satisfied with the simple functionality offered by ADS-B -
which would be fine if you *generally come into conflict with GA and
airliners more often than other gliders, but there are a bunch of us
for whom the opposite is true. Then the problem becomes some gliders
using Flarm and others using ADS-B, you lose some of the Flarm
benefits of path estimation for the non-Flarm gliders.

9B

9B

You are probably correct that no one is going to beat FLARM in an
optimized collision avoidance solution for high density glider
environments. *That's obviously their focus and they are good at it.

However, most recreational, non-contest pilots, primarily need a system
that will reliably alert them to other aircraft in their general
vicinity. *If I enter a thermal and know that there are 3 other aircraft
in the area, and I only see two, I'm going to abort and go elsewhere. *A
contest pilot obviously wants more data.

What is interesting about the Parowan situation is that this was not a
gaggle of gliders. *It was two gliders who apparently did not have a
proper appreciation that they were near each other. *A simple graphical
display that showed their relative positions, with a very simple
collision avoidance algorithm, or some form of auditory announcement
could have prevented this accident. *That's not to say that the FLARM
simulation was not impressive.

--
Mike Schumann

I played back the igc files from all the gliders flying that day and
can say that your speculation is not really supported by the facts.
Actually there were a number of gliders in that thermal - as it turns
out I passed right by it about a minute after the collision. There was
also a lot of non-thermalling traffic going in both directions at the
time, mostly within a pretty narrow altitude band within a thousand
feet or two of cloud base under a long cloud street.

An issue in these kinds of situations is that you can fixate on a
couple of gliders a bit higher in the thermal and miss the one
entering on a collision course with you at nearly the same time. A
cruder collision system has the potential to false alarm on too many
non-threats and on multiple gliders in the vicinity, making it hard
for the pilot to sort out which one is the real threat. Or it can
falsely identify a non-threat and mask the one that is really the
problem.

At this point I'm not at all sure why you'd pick straight ADS-B
(especially UAT) over something like PowerFlarm. The arguments keep
changing and hard as I try I can't find one that holds water when I
really run through all the issues. I think ADS-B in the long run is a
decent upgrade over PCAS, but PowerFlarm is more cost efficient and
more effective as a collision warning system, plus it has ADS-B in and
PCAS build in.

Also, I'll bet dollars to donuts that PowerFlarm gets FCC approval
well prior to ADS-B getting out from under the STC requirement.

9B

On 10/29/2010 4:11 AM, Andy wrote:
On Oct 28, 11:20 pm, Mike
wrote:
On 10/29/2010 12:51 AM, Andy wrote:





On Oct 28, 9:19 pm, wrote:

Also, to clarify, ADS-B does no path estimation of its own. That
function either would have to be added into an ADS-B unit by the OEM,
similar to the way Flarm does today - unlikely to be done in a glider-
specific way IMO - OR, it would have to be done by a separate external
device, perhaps a navigation computer/software like Oudie, WinPilot,
SN-10. For it to be effective manufacturers would all have to agree to
use the same algorithm, which also seems unlikely, unless they all
adopt the Flarm algorithm. That seems somewhat unlikely too, since I
don't think Flarm would want to start splintering how their algorithms
get used by splitting out the Flarm link technology from the collision
algorithm (which would have to be modified to accommodate the
differences in how path estimations get generated - with unpredictable
results). PLUS the external device OEM's would have to adapt to using
ADS-B inputs - another standards issue.

No matter how hard I try, it seems highly improbable that you will be
able to stitch together a satisfactory collision avoidance system for
gliders using ADS-B technology developed for general aviation. You'd
have to be satisfied with the simple functionality offered by ADS-B -
which would be fine if you generally come into conflict with GA and
airliners more often than other gliders, but there are a bunch of us
for whom the opposite is true. Then the problem becomes some gliders
using Flarm and others using ADS-B, you lose some of the Flarm
benefits of path estimation for the non-Flarm gliders.

9B

9B

You are probably correct that no one is going to beat FLARM in an
optimized collision avoidance solution for high density glider
environments. That's obviously their focus and they are good at it.

However, most recreational, non-contest pilots, primarily need a system
that will reliably alert them to other aircraft in their general
vicinity. If I enter a thermal and know that there are 3 other aircraft
in the area, and I only see two, I'm going to abort and go elsewhere. A
contest pilot obviously wants more data.

What is interesting about the Parowan situation is that this was not a
gaggle of gliders. It was two gliders who apparently did not have a
proper appreciation that they were near each other. A simple graphical
display that showed their relative positions, with a very simple
collision avoidance algorithm, or some form of auditory announcement
could have prevented this accident. That's not to say that the FLARM
simulation was not impressive.

--
Mike Schumann

I played back the igc files from all the gliders flying that day and
can say that your speculation is not really supported by the facts.
Actually there were a number of gliders in that thermal - as it turns
out I passed right by it about a minute after the collision. There was
also a lot of non-thermalling traffic going in both directions at the
time, mostly within a pretty narrow altitude band within a thousand
feet or two of cloud base under a long cloud street.

An issue in these kinds of situations is that you can fixate on a
couple of gliders a bit higher in the thermal and miss the one
entering on a collision course with you at nearly the same time. A
cruder collision system has the potential to false alarm on too many
non-threats and on multiple gliders in the vicinity, making it hard
for the pilot to sort out which one is the real threat. Or it can
falsely identify a non-threat and mask the one that is really the
problem.

At this point I'm not at all sure why you'd pick straight ADS-B
(especially UAT) over something like PowerFlarm. The arguments keep
changing and hard as I try I can't find one that holds water when I
really run through all the issues. I think ADS-B in the long run is a
decent upgrade over PCAS, but PowerFlarm is more cost efficient and
more effective as a collision warning system, plus it has ADS-B in and
PCAS build in.

Also, I'll bet dollars to donuts that PowerFlarm gets FCC approval
well prior to ADS-B getting out from under the STC requirement.

9B

Contrary to what everyone seems to think, I am not fundamentally opposed
to PowerFLARM. I understand the sophistication of its collision
avoidance logic, and it is very impressive and useful. I totally get
the necessity for reducing false alarms, so that the alarms that are
issued are meaningful.

My disappointment with PowerFLARM is the lack of a clear plan to take
advantage of the extensive ADS-B ground station infrastructure that will
cover much of the US in the next year or so, to provide the same level
of collision avoidance to transponder equipped GA and commercial traffic
that is available between PowerFLARM equipped gliders.

Granted, the PCAS capability built into PowerFLARM gives you some level
of protection, but you have no information on relative direction of the
threat, and a very crude estimate of its range. I find it very
difficult to understand how PowerFLARM will be able to suppress PCAS
initiated alarms from Mode C transponder equipped gliders in a gaggle,
while simultaneously still generating PCAS alarms from other GA aircraft
that also in the area.

The built-in 1090ES ADS-B In capability is great, but that doesn't
provide any ground station originated data unless you are transmitting
ADS-B out. The new Trig Mode S transponders provide this capability,
but require a GPS source. Is PowerFLARM going to provide this, or what
is the plan for glider pilots to end up with a complete ADS-B compatible
solution?

--
Mike Schumann

On Oct 29, 11:32*am, Mike Schumann
wrote:
On 10/29/2010 4:11 AM, Andy wrote:



On Oct 28, 11:20 pm, Mike
wrote:
On 10/29/2010 12:51 AM, Andy wrote:

On Oct 28, 9:19 pm, * *wrote:

Also, to clarify, ADS-B does no path estimation of its own. That
function either would have to be added into an ADS-B unit by the OEM,
similar to the way Flarm does today - unlikely to be done in a glider-
specific way IMO - OR, it would have to be done by a separate external
device, perhaps a navigation computer/software like Oudie, WinPilot,
SN-10. For it to be effective manufacturers would all have to agree to
use the same algorithm, which also seems unlikely, unless they all
adopt the Flarm algorithm. That seems somewhat unlikely too, since I
don't think Flarm would want to start splintering how their algorithms
get used by splitting out the Flarm link technology from the collision
algorithm (which would have to be modified to accommodate the
differences in how path estimations get generated - with unpredictable
results). PLUS the external device OEM's would have to adapt to using
ADS-B inputs - another standards issue.

No matter how hard I try, it seems highly improbable that you will be
able to stitch together a satisfactory collision avoidance system for
gliders using ADS-B technology developed for general aviation. You'd
have to be satisfied with the simple functionality offered by ADS-B -
which would be fine if you *generally come into conflict with GA and
airliners more often than other gliders, but there are a bunch of us
for whom the opposite is true. Then the problem becomes some gliders
using Flarm and others using ADS-B, you lose some of the Flarm
benefits of path estimation for the non-Flarm gliders.

9B

9B

You are probably correct that no one is going to beat FLARM in an
optimized collision avoidance solution for high density glider
environments. *That's obviously their focus and they are good at it.

However, most recreational, non-contest pilots, primarily need a system
that will reliably alert them to other aircraft in their general
vicinity. *If I enter a thermal and know that there are 3 other aircraft
in the area, and I only see two, I'm going to abort and go elsewhere. *A
contest pilot obviously wants more data.

What is interesting about the Parowan situation is that this was not a
gaggle of gliders. *It was two gliders who apparently did not have a
proper appreciation that they were near each other. *A simple graphical
display that showed their relative positions, with a very simple
collision avoidance algorithm, or some form of auditory announcement
could have prevented this accident. *That's not to say that the FLARM
simulation was not impressive.

--
Mike Schumann

I played back the igc files from all the gliders flying that day and
can say that your speculation is not really supported by the facts.
Actually there were a number of gliders in that thermal - as it turns
out I passed right by it about a minute after the collision. There was
also a lot of non-thermalling traffic going in both directions at the
time, mostly within a pretty narrow altitude band within a thousand
feet or two of cloud base under a long cloud street.

An issue in these kinds of situations is that you can fixate on a
couple of gliders a bit higher in the thermal and miss the one
entering on a collision course with you at nearly the same time. *A
cruder collision system has the potential to false alarm on too many
non-threats and on multiple gliders in the vicinity, making it hard
for the pilot to sort out which one is the real threat. Or it can
falsely identify a non-threat and mask the one that is really the
problem.

At this point I'm not at all sure why you'd pick straight ADS-B
(especially UAT) over something like PowerFlarm. The arguments keep
changing and hard as I try I can't find one that holds water when I
really run through all the issues. I think ADS-B in the long run is a
decent upgrade over PCAS, but PowerFlarm is more cost efficient and
more effective as a collision warning system, plus it has ADS-B in and
PCAS build in.

Also, I'll bet dollars to donuts that PowerFlarm gets FCC approval
well prior to ADS-B getting out from under the STC requirement.

9B

Contrary to what everyone seems to think, I am not fundamentally opposed
to PowerFLARM. *I understand the sophistication of its collision
avoidance logic, and it is very impressive and useful. *I totally get
the necessity for reducing false alarms, so that the alarms that are
issued are meaningful.

My disappointment with PowerFLARM is the lack of a clear plan to take
advantage of the extensive ADS-B ground station infrastructure that will
cover much of the US in the next year or so, to provide the same level
of collision avoidance to transponder equipped GA and commercial traffic
that is available between PowerFLARM equipped gliders.

Granted, the PCAS capability built into PowerFLARM gives you some level
of protection, but you have no information on relative direction of the
threat, and a very crude estimate of its range. *I find it very
difficult to understand how PowerFLARM will be able to suppress PCAS
initiated alarms from Mode C transponder equipped gliders in a gaggle,
while simultaneously still generating PCAS alarms from other GA aircraft
that also in the area.

The built-in 1090ES ADS-B In capability is great, but that doesn't
provide any ground station originated data unless you are transmitting
ADS-B out. *The new Trig Mode S transponders provide this capability,
but require a GPS source. *Is PowerFLARM going to provide this, or what
is the plan for glider pilots to end up with a complete ADS-B compatible
solution?

--
Mike Schumann

Somewhere earlier in this thread someone noted that they were planning
on doing that soon. I plan on doing it myself, except that the $4K
pricetag
means it will take a while to save up for it. Also, the TSO fiasco
throws
a big question mark over the ADS-B out part of the equation. My plane
(ASW-19) is type certified in the US, so the TSO requirement applies
to me.

-- Matt

Mike, I think you do not understand the fundamental difference between
ADS-B and FLARM.

ADS-B is designed for aircraft that do not want to fly close to each
other, and want/need long range situational awareness. Thats probably
where the whole ground station setup comes from - with all it's
disadvantages (coverage, latency, accuracy, etc.) It's a great system
for IFR traffic, and for GA flyers out sightseeing, but it sucks for
gliders in a gaggle or running a ridge.

FLARM, on the other hand, is specifically designed for aircraft
(gliders, helicopters) that often want to fly close to each other,
safely. In a gaggle, running a ridge, heading for the really nice Cu
near the glider field, on the Whites - gliders inherently will
congregate to find the best lift. That is the threat environment
FLARM is designed to cope with. And it's a proven item.

In a perfect world, there would be a SuperFLARM that would add some
sort of ADS-B in/out capability, so the benefits of both would be
present. But until then, FLARM addresses the immediate concern of
many, if not most glider pilots the best. A PowerFLARM, combined with
a mode s transponder, gives a glider most of the useful threat warning
capability of a full up ADS-B setup, and all the advantages of FLARM -
if FLARM is widely adopted by the US soaring community.

But if no-one gets ADS-b in their gliders (show of hands out there?),
then it's of no use in preventing the biggest historical threat -
glider on glider collisions.

I'm planning on getting a PowerFLARM next season to replace my PCAS.
I hope you will to, while waiting for your ADS-B to be installed.


Kirk
66

On 10/29/2010 11:37 AM, kirk.stant wrote:
Mike, I think you do not understand the fundamental difference between
ADS-B and FLARM.

ADS-B is designed for aircraft that do not want to fly close to each
other, and want/need long range situational awareness. Thats probably
where the whole ground station setup comes from - with all it's
disadvantages (coverage, latency, accuracy, etc.) It's a great system
for IFR traffic, and for GA flyers out sightseeing, but it sucks for
gliders in a gaggle or running a ridge.

FLARM, on the other hand, is specifically designed for aircraft
(gliders, helicopters) that often want to fly close to each other,
safely. In a gaggle, running a ridge, heading for the really nice Cu
near the glider field, on the Whites - gliders inherently will
congregate to find the best lift. That is the threat environment
FLARM is designed to cope with. And it's a proven item.

In a perfect world, there would be a SuperFLARM that would add some
sort of ADS-B in/out capability, so the benefits of both would be
present. But until then, FLARM addresses the immediate concern of
many, if not most glider pilots the best. A PowerFLARM, combined with
a mode s transponder, gives a glider most of the useful threat warning
capability of a full up ADS-B setup, and all the advantages of FLARM -
if FLARM is widely adopted by the US soaring community.

But if no-one gets ADS-b in their gliders (show of hands out there?),
then it's of no use in preventing the biggest historical threat -
glider on glider collisions.

I'm planning on getting a PowerFLARM next season to replace my PCAS.
I hope you will to, while waiting for your ADS-B to be installed.


Kirk
66

I think that you have a fundamental misunderstanding of ADS-B. ADS-B is
a general purpose system that is designed to permit aircraft to announce
their position and velocity vectors to anyone who is interested. For
all practical purposes, it has the same accuracy, latency, etc. of FLARM
based systems. It doesn't make any difference if aircraft are flying in
close proximity or not.

FLARM does the same thing and then adds another layer of collision
detection logic on top of that, which is specifically optimized for the
glider world. There is no reason that the same logic could not also be
added on-top of an ADS-B based position sensing technology. In fact,
PowerFLARM claims to be doing this using their 1090ES IN capability, if
the user has ADS-B Out installed.

The ground station ADS-B infrastructure is being deployed for 4 reasons:

1. To provide translation of ADS-B transmissions between UAT and 1090ES
(an unfortunate necessity due to the FAA's decision to go with 2 ADS-B
formats in the US)

2. To transmit TIS-B data. This lets ADS-B IN equipped aircraft see
the position and altitude of Mode C and S Transponder equipped aircraft
which are not yet outfitted with ADS-B transmitters.

3. To transmit additional weather and NOTAM information.

4. To detect ADS-B equipped aircraft so that they can be displayed on
ATC radar displays.


--
Mike Schumann

Mike, you have just confirmed what I just suspected. Again, ADS-b is
designed for keeping airplanes that want to stay apart, apart. And
yes, I understand exactly what the ground infrastructure is for, and
it's not gliders - It's for IFR air traffic control and GA planes
working their way through bad weather, checking on the latest NOTAMs
at their destination. Sure, "someone" could develop software and
hardware to use ADS-b to do what FLARM has been doing for the past 10
years, BUT THE CHANCE OF THAT HAPPENING IN THE NEAR FUTURE IS A CLOSE
APPROXIMATION OF ZERO!

Can't you understand that this isn't a zero sum game? NOTHING
prevents you from having FLARM now (in Europe) or next year (in the
US, hopefully) and later, if and when ADS-b becomes affordable and
useable in a glider cockpit, installing ADS-b in their glider. And
that even if the FAA gave a free UAT to every aircraft out there,
FLARM would still be useful in the glider community?

The two systems overlap each other, but occupy different requirement
niches.


Kirk
66

On 10/29/2010 12:33 PM, kirk.stant wrote:
Mike, you have just confirmed what I just suspected. Again, ADS-b is
designed for keeping airplanes that want to stay apart, apart. And
yes, I understand exactly what the ground infrastructure is for, and
it's not gliders - It's for IFR air traffic control and GA planes
working their way through bad weather, checking on the latest NOTAMs
at their destination. Sure, "someone" could develop software and
hardware to use ADS-b to do what FLARM has been doing for the past 10
years, BUT THE CHANCE OF THAT HAPPENING IN THE NEAR FUTURE IS A CLOSE
APPROXIMATION OF ZERO!

Can't you understand that this isn't a zero sum game? NOTHING
prevents you from having FLARM now (in Europe) or next year (in the
US, hopefully) and later, if and when ADS-b becomes affordable and
useable in a glider cockpit, installing ADS-b in their glider. And
that even if the FAA gave a free UAT to every aircraft out there,
FLARM would still be useful in the glider community?

The two systems overlap each other, but occupy different requirement
niches.


Kirk
66

The ground infrastructure is for EVERYONE! Glider pilots are probably
not as interested in weather or NOTAMs. They are interested in seeing
Mode C / S equipped aircraft (both GA and Jets) in their vicinity. This
is the BIG reason you should care about ADS-B ground stations.

The PowerFLARM proponents claim that it will handled 1090ES ADS-B
Inputs. Does this include TIS-B data? What is the plan for PowerFLARM
equipped aircraft to transmit ADS-B Out data so that the TIS-B data is
visible? If PowerFLARM can do that, then it will be a killer product,
not just for the glider world, but also for GA in the US.

--
Mike Schumann