If the AAD fired but too low because of turning it on at an off altitude? Missfire at high altitude? Not fireing at all (yes)? Missfire in airplane? If it fired correct but person backflying causing delay? Fire due to low turn?
(This post was edited by Hellis on Oct 11, 2010, 8:23 AM)
You'd have to pay attention to how they define "save" to see if the numbers are comparable.
And how would you define "fail?" The AAD's job description isn't to save someone's life -- no machine can be designed to do that.
The AAD is a machine that responds to physically-detected stimuli; primarily the rate at which air pressure is changing, as compared with a baseline set by initializing it at ground level. If you initialize it at the wrong place, it can "fail." If the barometric pressure changes significantly, it can "fail" (although that's low on the likely tree). If the AAD happens to be in a burble because of the jumper's body position, it can "fail." If the reserve chooses that day to have a pilot chute hesitation, it can "fail." If there is debris blocking the sensing unit (unlikely, but hey), it can "fail."
But none of these are failures of the unit; they are all violations of its operating parameters. A person is required in the loop, to make the complex decisions needed for skydiving; a machine that could make all of those decisions would be incredibly more complex, more expensive, and have far more (and more sophisticated) failure modes.
There are quite a few different AAD's out there. They are a back up, nothing else. The primary AAD is the jumper. The AAD should be irrelevant on every skydive, and if you depend on one for survival, then you need to quit skydiving now.
There is often no reliable data to say exactly when an AAD has activated, or in fact whether it was relevant to events that occurred.
After a fatal accident the unit is often damaged to the point of making accurate analysis impossible. Any relevant data about the AAD would be contained within the fatality report, which is a public document.
Having said all that, AAD's when used properly, are very reliable. If they weren't, they would be a big liability to safety, and jumpers would not use them.
Jumping an unreliable AAD would be Russian roulette because it would kill you in the end.
(This post was edited by obelixtim on Oct 11, 2010, 8:59 AM)
My post is implying that "failure" is pretty narrowly defined. I don't know if results are made public, but in the US at least, with all of the litigation running around, and people who want to define "failure" as "little Buffy died and she thought the AAD would save her" I wouldn't be surprised if they stay private.
AADs are tools. They're pretty good, but if you want to significantly reduce "failures" then you might have to significantly increase times when they fire unexpectedly. That's less acceptable than a failure to fire when someone thought it should. Cypres's reliability in that sense if exactly why AADs only became common after it came on the market.
My post is trying to say that the failure parameters of an AAD are very rarely clearcut, and a good indication that people should open their own parachutes.
>personally, I'd like to see an unbiased/3rd party report of AAD statistics.
Unfortunately that is next to impossible. The one person with firsthand experience with what happened will nearly always be biased, since there are generally penalties for being 100% honest about what just happened.
Not even the AAD manufacturers have complete data on their products. Remember, some AAD's don't need new parts when they activate. And for those that do often the information is just not supplied to the manufacturer, let alone a third party.
They work most of the time. If it was SUPPOSED to fire and didn't or the reserve didn't open in time , well you were dead anyway. If it fires and it wasn't supposed to, either its bad or its just a pain. But it's the cost of wearing one.
There were two big arguments against AAD's before the CYPRES; one they're two big/heavy and two they aren't reliable enough in terms of firing when they aren't supposed to.
CYPRES came out and the arguments were price and still "I don't want something that can open my reserve when I don't want it to." The first one is always there. The second one turns out to be rare. We all expected newer AAD's to kill someone eventually.
Pre CYPRES we had a friend bounce after being knocked unconcious exiting on a DZ 20 some way at the 1985 Freek Brothers. Lots for folks said they were going to get an AOD (the acronym at that time) or at least start jumping a hard helmet. Not ONE AOD or helmet was bought because of that. We knew/know the risks and accepted them.
IT'S SKYDIVING. Before AAD's you were DEAD once you left the airplane unless you took a POSITIVE action. Few situations in life are so certain. With AAD's you now have a chance to live. But no guarentee. Just read your parachute owner's manual! If you want a guarentee your not going to die skydiving stay on the ground.
Imagine what the fatality statistics would look like if we had ADDED all the canopy piloting deaths TO the bounces instead of SUBSTITUTING them. AAD's starting saving lives when people starting killing themselves under open parachutes.
Very good answer Wendy. When there is a problem with an AAD it's not that simple to determine what has happened. There is so many factors contributing to a firing. Too many people react emotionally and are prompt to blame such and such company without knowing instead of trying to stay calm and get a well written technical report. I know that a complete report takes time for having had myself to do an expertise once in a case of a fatal accident. Expect also to never get a certitude about what happened. This is not only right for AADs but also right for airplane accidents. But there is a way to know more about what happened: The only company which provides an interface to customer to download figures and diagram altitude-time of the last 16 minutes recorded on an AAD is Vigil. When an AAD fires or anything "wrong" happens in the air, it is very interesting to see a diagram of what has happened. On the altitude-time chart, one can see dot of different colors for start of the dive, opening altitude and firing within fraction of a second including altitude. On a Vigil a red dot is the start of the dive, the yellow dot is the opening and the blue dot is the firing if any. The chart or diagram shown in the attachment illustrates an actual firing of a Vigil 1 at Perris Valley in 2005. The firing was due to a low pull, too many people above me at the separation (not an excuse ) (no. 17 on the Vigil web site). The firing occurred at 1099 ft (1100ft) since I was in an upward position while my main parachute was beginning to deploy. Therefore there was no differential pressure between chest and back equivalent to 260ft. This is why the Vigil 1 fired at that altitude (840 ft is the firing altitude when flat on your stomach) and in an upward position it is 840 ft + 260 ft = 1100 ft. The chart shows that if I had pulled a fraction of a second sooner, the firing would have been inhibited. The result : a beautiful biplane which I landed without any problem. The green curve is the altitude-time curve. The yellow curve is the speed-time curve. Even if the speed curve is referring to sudden changes of pressure, the AAD software does a sampling and rounds up the data to "decide" if the firing speed is met or not. This diagram was available using my software and interface provided by Vigil as an option. This diagram therefore wasn't done by Vigil. I got it from a computer. Protrack can also provide the same type of chart if using their software and interface. Since on Dropzone.com we are restricted at 200k picture, the picture in attachment was the best I could provide. Note: I am also using the capability of the Vigil interface and software for using the chart to determine how long it takes for my main to open. Sabre 2-170 340 ft (average on 10 jumps) Katana 170 600 ft (average on 10 jumps)
(This post was edited by erdnarob on Oct 21, 2010, 12:00 PM)
up to date: Cypres saved over 2000 Vigil saved 81 Argus saved 22
I don't like the word 'saved' in this context. These were AAD fires - many of them were not 'saves'. Otherwise the figures suggest that an average of 110 people would have been killed annually by no-pulls in the 19 years since the Cypres was first available - and presumably most of those 2,100 fires have actually occurred since AAD use became the norm, which is a significantly shorter period of time.
An average of 110 bounces through failure to pull a year is unlikely, given the rate back when AADs were only used by students (or even early freefall students).
We have nearly eliminated the category, but it still usually never made up more than about 30% of the deaths. And we have invented high performance approaches (especially in traffic) to help take up the slack.
Meaning not at all to argue, but only to continue to discuss -
Who decided if it is a "save" or not?
If these 2000 CYPRES saves were "saves" according to the jumper who got "saved", maybe it does mean that 2000 more people would have died over the period in question.
If someone reported it as a "save", maybe it really was.
Do you think that every single activation is reported to Airtec and they call them all "saves" even if they were not?
I suspect that not all activations are reported, which might make the 2000 number actually be lower than it really is.
The nature of the jumping population now is pretty different from the nature of the jumping population 25 years ago, so I wonder how meaningful it is to compare the "no pull" rates from then to the "save" rates from now. I know lots of people now who simply won't jump without an AAD. Back then, everybody had to have the confidence that they could save their own life. Now it is not so clear as that. So, as format says, maybe a lot of these "saves" wouldn't have been jumping at all were it not for the AAD.
But, they are jumping, and if they are reporting these events as "saves" maybe they really are.