On Sunday, January 10, 2016 at 6:45:59 AM UTC-8, Greg Delp wrote:
On Sunday, January 10, 2016 at 5:42:30 AM UTC-5, XC wrote:
On Sunday, January 10, 2016 at 12:57:03 AM UTC-5, smfidler wrote:
XC,

Again, you assume the flawless function of the Flarm system. 25 seconds, if the Flarm "network" is not perfectly functional due to say structural interference, may be reduced to 5 seconds. Or zero. Arbitrary discussions about were "safety ends" and "philosophical Flarm fairness?" (whatever the hell this should be) begins has proven quite frightening to observe.

The longer the time period the Flarm has to make a potential warning (beep as "bogey" appears), the better. This is a fact (from a safety viewpoint). Artificial reductions in Flarms engineered performance potential will by definition also increase the odds of a dangerous situation occurring. While being far better than nothing, the Flarm network is still prone to regular coverage challenges and clearly does not have, perfect, 100%, 360 degree coverage at all times. Far, FAR from it.

We need to kill this whole idea for at least one year. It's out of control. It is irresponsible (at best) to screw around with Flarms potential range at this point. The unintended consequences are potentially huge. The philosophical competitive "fairness" argument is simply no peer to safety. It's not even on the same planet as safety. Stop treating this discussion like this as a debate between equals. Start respecting the fact that Flarm, while valuable for safety, is far, far from perfect or infallible. So when you say a supposed time value that you find acceptable for your little crusade, imagine that the gliders 25 seconds away from a potential collision are not seeing each others antenna at 25 seconds for whatever reason. Or maybe you get a proximity "beep" and then lose coverage for 15 seconds..

Bottom line, some here care far more about philosophical fairness arguments than the significant safety value Flarm provides via its ability to create situational awareness for us. This crusade become an literal obsession for some.

I won't debate or compromise Flarm at this point until some sort of impartial, objective study (and extensive testing) has been completed. This was the requirement of my initial support for a limitation. Since then, the almost reckless abandon that many supporters (almost certainly coordinated in a pseudo RAS Flarm "mode" campaign) have demonstrated here is a real problem. The USRC opinion poll does not support this desperate position.

Perhaps the FAA should be involved. Has anyone discussed this topic with the FAA? I wonder what their take would be.


I was just pointing out some of the time values involved in operating with TCAS. I suspect that research was more extensive and the technology is more proven. Many TCAS resolution advisories occur before they are displayed as targets on the screen. The pilot's eyes are directed to the display (on the VSI in our case) to make a correction because it is a system used during IFR.

With FLARM the eyes are directed outside as they should be. The best way to do this is through the audio warning system. More could be done with this to improve safety. For example, there could be a warning "Multiple traffic, One o'clock high." This is another real suggestion that would enhance the FLARM system. The TCAS time values are real numbers to add perspective to the discussion.

XC

Then point out all the time values not just the ones that seem to support your stance on stealth. TCAS Non threat traffic more than 6 miles away can be displayed to well over 40 miles away. Proximity intruder traffic less than 6 miles away fill in the diamond display(visual no audio) to get your attention and provide situation awareness to prevent surprise TA and RA alerts. Traffic Alerts (TA) for potential collisions will happen from 20-48 seconds to closest point of approach(collision) with a visual change to yellow diamond and audio "Traffic Traffic" Under normal conditions a TA will proceed an RA by 15 seconds which prepares the pilots for any required RA avoidance maneuver while giving time to acquire the target visually if possible. Resolution Advisories (RA) which TCAS deems an imminent collision threat with a time of closest approach of 15-35 seconds warns with a red diamond visual cue and audio avoidance commands (Climb Climb Climb, Monitor Vertical speed etc...) as well as the required vertical speed required for the 5 second maneuver you posted.


TCAS uses these multiple steps of awareness from well over 40 miles away to try and prevent surprise RAs as well as to give crews the time required to scan visually for the traffic. I also find it very hard to believe there have been many instant RAs without the targets going through the normal threat scales of non threat, proximity, TA, RA, Clear of Conflict or at least some of them. Does it happen yes but it is not the norm.

To suggest that your eyes are inside looking at the TCAS only because you are IFR is misleading. It gives you the fastest way of locating the direction and altitude of the potential collision traffic so you can then start looking visually for the traffic if your not in clouds. No one using TCAS gets a pop up TA and starts randomly visually scanning the sky hoping to see the traffic. It takes much longer than looking at the TCAS visual display for a split second to find out where the traffic is then back outside to find it visually.

This is exactly the same method that FLARM uses with a scaled response to traffic threats as they get closer and are deemed a threat. Both TCAS and FLARM allow much better situational awareness of traffic. I would argue that FLARM stealth mode is exactly like the pop up surprise TCAS RA example you used. It should never happen as there is only 5 seconds to respond. TCAS users get the advantage of a computer giving an immediately calculated escape maneuver. FLARM users only have seconds to find the traffic visually to find out what to do.


There has been lots of good work on human reaction times (much of it as relates to OODA-loop types of cognitive processing). If you also look at the research studies done on visual scan and collision avoidance, you realize that there are some profound differences between how Flarm and TCAS operate with the pilot(s) in the loop. An RA climb/decend command is pretty unambiguous and the context is pretty clear - push or pull on the stick as commanded or you will hit someone. You can act with confidence that the systems is giving complementary commands if both aircraft are TCAS II equipped. The "Orient" phase of this "Observe-Orient-Decide-Act" loop in collision avoidance has the biggest potential variance in time because it is not always 100% clear what is happening based on limited inputs. An RA cuts this time down dramatically, just Act on what the system tells you. Flarm doesn't do this so the pilot needs to go through an entire OODA-loop cycle to figure out what to do.

I'd also call into question the wisdom and effectiveness of trying to implement any strategy that is 100% eyes out the window with audio cues only. Audio cues take a long time (multiple seconds) to generate precise information on distance, relative altitude, track, speed, orientation, yet much of this information can be important in the "Orient-Decide" part of the loop. Picking up a collision target visually is hard (as opposed to one that is not a threat and therefore has angular movement in the pilot's visual field). The pilot has to orient his gaze to the right location +/- about 2 degrees to pick up a target - peripheral vision isn't much help. With enough scanning around you may pick up the target, but getting into the right ballpark more precisely with the aid of a look at the display may in fact be a faster and more effective way to acquire a threat target visually.

I don't think TCAS is intended for RAs to be the first line of defense, rather, they are the last line. Being able to have most of the "Observe-Orient" part of collision avoidance mostly taken care of with traffic 5+ miles away is preferred, especially since most of the preemptive maneuvering will be with an Air Traffic Controller in the loop for TCAS-equipped traffic. Flarm and gliders instead of SSR/TCAS and jets present very different scenarios so direct comparisons are challenging, especially if you are talking about timing.

9B