FLARM in Stealth Mode at US 15M/Standard Nationals - Loved It!

Andy-

I think you mean well and speak heartfelt words (along with the whole group here)...but I also feel most of this discussion is highly subjective. I would like to see at least some data or proof.

My questions were serious and (hopefully) meaningful. I did not mean to throw you under the bus. I was hoping to get some more information on the use of FLARM at the world level. I think that conversation may be a useful one here as it is more objective (assigned tasks, smaller turn areas, more even pilots, etc).

In the U.S., I honestly don't believe that FLARM has the slightest effect on the results. Some may try their hardest, but it's just to hard to do consistently. Also, FLARM data is the norm at the World Championship level and all but one U.S. Contest in all history. If we run off and adopt "stealth mode," we once again handicap our world level pilots by "watering down" the game.

Perhaps the elite, sophisticated world gliding teams (Germany, Poland, Britian, etc) would be able to do some damage with FLARM data. But good on them because the safety based FLARM technology is equally open to all parties, so it's a level playing field. Again, I see FLARM data as small part of the overall equation. As long as it's level for all, no problem. This is my view.

That said, I have recently heard "campfire stories" of very sophisticated, coordinated use of FLARM at world championships. "Hard to believe" stuff. As it is third hand and I was not there, I can't post it here, sorry. I have also heard the FAI is considering stealth mode requirements. So maybe there is real data out there on how this is changed results, etc. I was hoping to hear more about this and that was the reason for my questions.

I just don't hear any specifics here...I hear broad statements and assumptions.

Sorry to make you feel pinned down like that. Not my intention.

I really enjoyed meeting you at Hobbs and think you have a very bright future in soaring. I look forward to flying with you again soon!

Sean

On Friday, August 7, 2015 at 12:53:32 PM UTC-4, Sean Fidler wrote:
Andy-

I think you mean well and speak heartfelt words (along with the whole group here)...but I also feel most of this discussion is highly subjective. I would like to see at least some data or proof.

My questions were serious and (hopefully) meaningful. I did not mean to throw you under the bus. I was hoping to get some more information on the use of FLARM at the world level. I think that conversation may be a useful one here as it is more objective (assigned tasks, smaller turn areas, more even pilots, etc).

In the U.S., I honestly don't believe that FLARM has the slightest effect on the results. Some may try their hardest, but it's just to hard to do consistently. Also, FLARM data is the norm at the World Championship level and all but one U.S. Contest in all history. If we run off and adopt "stealth mode," we once again handicap our world level pilots by "watering down" the game.

Perhaps the elite, sophisticated world gliding teams (Germany, Poland, Britian, etc) would be able to do some damage with FLARM data. But good on them because the safety based FLARM technology is equally open to all parties, so it's a level playing field. Again, I see FLARM data as small part of the overall equation. As long as it's level for all, no problem. This is my view.

That said, I have recently heard "campfire stories" of very sophisticated, coordinated use of FLARM at world championships. "Hard to believe" stuff.. As it is third hand and I was not there, I can't post it here, sorry. I have also heard the FAI is considering stealth mode requirements. So maybe there is real data out there on how this is changed results, etc. I was hoping to hear more about this and that was the reason for my questions.

I just don't hear any specifics here...I hear broad statements and assumptions.

Sorry to make you feel pinned down like that. Not my intention.

I really enjoyed meeting you at Hobbs and think you have a very bright future in soaring. I look forward to flying with you again soon!

Sean

hey sean,

i didn't feel thrown under the bus or pinned down. i was happy to elaborate on my thoughts, since i made a pretty short statement that was contradictory to what i had previously written. this discussion is interesting to me, and i'm glad to take part in it/read everyone else's opinions. on that note, I agree that it IS a subjective discussion.

sadly i can't give you more first hand knowledge of flying with flarm at the world level... YET look forward to flying with you again too!

ND

On Friday, August 7, 2015 at 9:53:32 AM UTC-7, Sean Fidler wrote:

I just don't hear any specifics here...I hear broad statements and assumptions.


One last effort to put the data we have on the table. I removed Days 1 and 6 (mass landout days) from Harris Hill. This leaves only days that had high completion ratios - only P7, XC and BT flew both 2014 and 2015 15-Meter Nationals and could make informed judgements about how "randomizing" the conditions were for what remained (outside the two landout days, that is).

The result is Harris Hill had an average winners score of 952 versus 869 for Montague, so from a devaluation perspective dropping the two mass landout days makes Harris Hill quantitatively less random. Even so, pilot performance averaged 10% off of what you'd expect from the PRL versus 7% for Montague (and 6.5% if you exclude P7's "random" landout).

I'd add that at Montague I saw lots of pretty complete Flarm setups (I worked on a bunch - doing config files, RF testing and even installed my spare unit in a competitor's glider - which is why people kept seeing 9B in multiple places on course - heh heh). I'd have to say that I think many, if not most, pilots had the ability to use Flarm tactically - I certainly did when I could and there was even a protest based in part on use of Flarm data for tactical purposes. Flarm may not have been in use by everyone to 100% of its ultimate capability, but it was certainly in use by enough people that if it was making a big difference you ought to be able to detect something. HH randomness notwithstanding, the Montague results were pretty orderly - so Flarm leeches were not successfully crawling up the scoresheet - though some may have tried.

The data to-date provides no evidence that people are generating sustained benefit from tactical Flarm usage. In fact, the evidence is that people are NOT generating a sustained benefit. Maybe that will change with time and experience.

This discussion has raised another thought. The single biggest factor in randomizing scores away from what makes sense (at least based on the long term competition records of the pilots) is weather. It has been argued here that it is harder to pick up any signal from Flarm leeching with random weather noise in the scores from tricky days that land out most of the field except for a lucky few. So, should we be disallowing days where less than 2/3 or 3/4 of the fleet finishes? They say contests are won on the weak days, maybe that's bad thing?

I tend to agree with Sean (7T). There seems to be a lot of speculation and supposition and "anecdotalytics" floating about - on all sides. It does appear to me that Stealth mode retains enough warning range for most collision scenarios, but how restricting the broader situational view might cascade into a threat scenario under particular circumstances is unclear and I would want to know a lot more about that. For instance, I have already once mistaken a glider that didn't show up on Flarm for one that was a threat and mistakenly tracked the wrong target until way too late. I suspect trial lawyers don't care if Stealth is 99% as good as no Stealth in the event anything bad happens. That's a concern.

I also tend to agree with 7T in his view that tactical Flarm usage sometimes helps a bit, sometimes hurts a bit. In FAI-style team and gaggle flying it probably can be used to greater benefit as gaggle-hopping is a strategy that can win races (Sebastian Kawa has a view on this point). Gaggle-hopping is a strategy that may be enhanced by, but was not created by, Flarm. It's more a result of FAI scoring philosophy. In any event, the strategy is available to everyone. Also, with or without Flarm, flying more assigned tasks will likely generate more leeching behavior, so if you're opposed to leeching, you might not like ATs either.

I also don't buy the "expensive arms race" argument. Most US racing pilots have Flarm available to them today (if you're really poor I'll loan you my spare, or you can rent one). Most glider pilots have some sort of moving map display - whether it's an LX9000/ClearNav premium computer or an Oudie (basically the same software as the LX) or a smartphone running open source gliding software. They all have Flarm tracking basically for free and new Flarm features will be made available to the broad market (at least that's been the case so far). Turning features off won't make the devices or software any cheaper and they won't not develop them just because racing pilots don't like them - they're great for OLC/buddy flying.

That does leave us with what seems to be the essential point. Does Flarm increase or decrease (some might say ruin) the qualitative experience of glider racing? There's no better feeling than having a strategic inspiration, going off on your own and smoking the field. There's also something disheartening about struggling down low until you find the boomer that saves your day only to have the gaggle roll in right on top of you. Stories of pilots who leeched their way to the podium by following top pilots around have been part of soaring folklore as long as I can remember, but I don't generally think the people on the podium are second-rate and most scoresheets make sense to me most of the time. BUT, there is also GW's sentiment - and it does not seem so rare - that Flarm allows pilots to keep in touch with some of the field and generates more of a sense of racing and some camaraderie in a sport that can be isolating. He argues that that's more fun and makes the sport more accessible to new pilots, even if it doesn't catapult them to the top - or even middle - of the scoresheet. We struggle with limiting the shrinkage of the sport. I'd be curious to know whether newer racing pilots learn more quickly and enjoy the experience more when they can see more of what's going on out of course.

Good discussion - many useful points and perspectives.

9B

So, for the sake of argument, let's say that Flarm leeching is no factor in contest results. Does anyone think that means tactical use of Flarm is not resulting in more time with eyes in the cockpit? Seems to me that the supposed safety benefit of not being in stealth mode may be negated by pilots fixated on their screens.

So, for the sake of argument, let's say that Flarm leeching is no factor in contest results. Does anyone think that means tactical use of Flarm is not resulting in more time with eyes in the cockpit? Seems to me that the supposed safety benefit of not being in stealth mode may be negated by pilots fixated on their screens.

I agree - lots of good discussion.

I would again emphasize though that nobody has yet put the effort into building a true Tactical Leeching Tool. Flarm just provides data and a collision warning algorithm which seems to work extremely well for the purpose for which it was designed. I have been truly impressed at how few spurious warnings I have gotten in either unrestricted or stealth mode. As a safety tool, I love it.

But, please consider what folks could do IF they put their minds to it. Visualizing and presenting the Flarm data for tactical advantage has hardly been tapped at all. Imagine a filtering algorithm that takes in Flarm data from a bunch of gliders. It figures out if this is a "gaggle" (defined by at least 2 gliders circling for at least 30 seconds and climbing, for example). Now, imagine that it smooths the climb data from the gliders that are identified as Gaggle A with an indication of average climb rate (30 second, 60 second, duration of climb) as selected by the user. Imagine that another gaggle (Gaggle B) nearby is similarly filtered and displayed. Color coding is applied with strongest average climbs in Green and decreasing in size and intensity with weaker climbs. Also, the user can set alerts to indicate if specific targets are in Gaggle A or Gaggle B. Okay, Gaggle A is clearly averaging a better climb, and as a bonus P7 is part of the group. This is not 10 years out sort of stuff. I've already written specs for several of these examples.

I suspect that the reason a lot of folks are finding Flarm of limited tactical value right now is largely unfamiliarity with the tool combined with User Interfaces which are not yet optimized for competition. My Flarm routinely provides 7.5km to 8km range, which is basically a corridor 15km wide (left and right). And contrary to statements made in this thread, many thermals are relatively persistent and regenerative. All you have to do to see this in action is to download competition files from a recent contest and replay the flights in maggot race mode on SeeYou. If you synchronize on start time, you can see gliders converging on exactly the same spot often times 10-15 minutes apart when you shift back to real-time. So, having a choice of 2 gaggles displayed each of which is only 3 miles away and KNOWING which gaggle is the stronger climb is a very real tactical advantage. Similar examples apply to the pre-start, straight glides, etc.

Of course, if everyone has it, then maybe the playing field is again level. Except that instead of looking out the window and eyeballing other gliders to figure out which group is climbing better relative to you, you'll be zooming in the gain on your Leeching Window. Sort of like good weather radar operators today who know how to identify the really ugly cells vs. the more benign cells. Maybe that's the exciting next frontier?

P3

So we should impose stealth mode now because somebody might some day write some killer software that might let people know where thermals are and this might turn out to be a bad thing? I've "written specs" for lots of stuff too, like thermal detectors. No reality yet.

I think we need to get back to a simple principle: Let's see if something is really a problem before we start passing rules against it.

Surely, you guys who want to impose stealth mode can come up with some real, serious, documented problem that real flarms today are causing, not just hypothetical problems of hypothetical future software?

By then we'll all have FAA mandated ADSB displays of all traffic, super cheap drone anti-collision technology showing us where the thermaling birds are, and so on.

John Cochrane

The cadence of transition remains ever present, clocks that bind will be left to rust. Tick tock, tick tock.....

The old days are in the past. This argument is similar to the GPS data loggers versus cameras. The cream always raises to the top. Why limit Flarm. No Flarm data is going to help you out climb, core faster, or fly more efficient than a better pilot.

On Friday, August 7, 2015 at 7:49:31 PM UTC-7, John Cochrane wrote:
So we should impose stealth mode now because somebody might some day write some killer software that might let people know where thermals are and this might turn out to be a bad thing? I've "written specs" for lots of stuff too, like thermal detectors. No reality yet.

I think we need to get back to a simple principle: Let's see if something is really a problem before we start passing rules against it.

Surely, you guys who want to impose stealth mode can come up with some real, serious, documented problem that real flarms today are causing, not just hypothetical problems of hypothetical future software?

By then we'll all have FAA mandated ADSB displays of all traffic, super cheap drone anti-collision technology showing us where the thermaling birds are, and so on.

John Cochrane

Well, I am no Luddite, but will there be never be a technology that we ban from our cockpits? Won't we have to draw a line at some point?

I hated turn point cameras and I really would not want to do without GPS, but before all the gadgetry, contests usually filled up or were oversubscribed.

Surely, you guys who want to impose stealth mode can come up with some real, serious, documented problem that real flarms today are causing, not just hypothetical problems of hypothetical future software?

John Cochrane

John,

The problem is here. FLARM is a great help in finding the best competitors pre-start, following them out the start gate, and good way down the course line. For a short task in weak conditions folks can follow the top guy all the way around. XG never had a chance of hiding out and sneaking away.

This results in artificially similar flights and scores, not reflective of the variety of ability in the race, and make for a duller contest.

Here is a video replay of PAGC Day 3. Let me say that I and my partner would have gladly helped wear a groove around the course line. Those were the rules at this contest and it was clearly the fastest way to get around. 5E had a late start and got stuck low so I waited for him. It cost us.

https://youtu.be/Oti1iLPcTKg