Which has a higher rate of failure a SPG or a transmitter?

[beaverdivers]

From what beaverdiver has described, that is not what RBT tells you. RBT assumes you will continue consuming gas at your current rate and tells you how much time you have with no gas consumption rate changes before you must start an ascent. .

I not sure where you saw I described the above false statement.

 

[eponym]

I would be extremely skeptical of any statement made by a dive shop owner/employee which asserts that analog SPGs have a higher failure rate than wireless AI transmitters. My skepticism is based on the following:

1. Anecdotal evidence can be misleading.
2. In the absence of true failure rates (total number of failures/total number of uses), a fair comparison regarding reliability really can't be made.
3. The dive shop stands to make more profit per unit sold on a wireless AI computer system than an analog SPG + non-AI computer (or analog SPG + analog depth gauge + timing device).

Any dive shop that claims it sells wireless AI computer systems vs. SPGs in a 10:1 ratio must be using highly manipulative sales tactics with its customers. There is a high likelihood that it is providing its customers with inaccurate and/or misleading information regarding the failure rates of the respective pressure-sensing devices. I don't think that's a dive shop I'd want to do business with.

Analog SPGs are relatively inexpensive (compared to electronic AI computers). A spare SPG + HP hose can be easily carried in a save-a-dive kit. When analog SPGs "break," they are almost always replaced not repaired.

Although hoseless AI computers can make gas checks very convenient during the dive, I have been on a fair number of dives where a diver was cussing at his hoseless AI computer during pre-dive checks because it would not sync up with the transmitter. If I had a hoseless AI computer, I would always carry an analog SPG + HP hose in my save-a-dive kit.

Good observations, especially on frequency of use, which would be the denominator in failure statistics.

And I suppose it's fair to assume that a retailer has some selective bias.

 

[redacted]

It was worth posting twice!

We use the Remaining Bottom Time ( RBT ) in our beginning pool session to see how comfortable a NEVER EVER student is. Then we have them watch their RBT through out their training ( pool & OW ).

It is fun to see their faces when they see their RBT go from 30 mins. ( while they are swimming ) to 60 mins. ( while they are resting or doing a fin pivot ). You could use a SPG and a watch, but the Galileo is taking a sample every 4 sec.

After each dive we download & debrief from both the pool & OW.

A SPG can NOT do that!

We have been using this teaching method since the late 90's.

Now, the Galileo Sol enables us to incorporate the heart rate!

Perhaps I misinterpreted your post above. It seems to be saying that SAC is being estimated by the computer over short period of time and that SAC is used to forecast gas usage in the immediate future. Is that incorrect?

I not sure where you saw I described the above false statement.

Unless it is substantially incorrect, where is my false statement?

 

[DaleC]

The Air Integrated computer ( AIC ) is safer. I just started a new thread for this discussion.

This thread is addressing failure rates of a SPG & transmitters.

AIC's can use a High Pressure ( HP ) hose
or a transmitter. AIC's become hoseless when getting rid of the HP hose.

Question: Do you use a remote for your TV?
Yes. But I have cursed many times when I cannot find it and the TV in question has no TV based controls - it becomes effectively a large paper weight. But to return the question: Is a remote a want or a need?

Do you use a garage door opener?
No, I have a garage door but lift it manually.

Do you have a cell phone?
Yes, but it doesn't even take pictures and compared top the latest generation of smart phone it is stoneaged. However, I still manage to do everything I need to do with it - which... is phone people.

Can you even find a phone booth? If you are under the age of 20, do you know what a phone booth is?
Interesting. Are you suggesting that SPG's will be the phone booth's of tomorrow? If so, would that be a good thing?

Do you use a J valve?
Why yes I do!

Answers in quote above.

Eurodog, No thanks.

 

[Paladin]

I have several SPGs that were made in the 1970s and I have never had a single SPG failure. And, like DaleC, I also use J-valves.

 

[halocline]

Don't worry, beaverdivers, as soon as the next, even more expensive and excessive SP dive computer comes out, you'll be back to tell us how diving with that antiquated Galileo sol is dangerous and just plain no fun.

Until then, I'm not even getting in the bathtub without HAL.

Me: HAL, how much dive time do I have left?
HAL: I'm sorry Dave, I just can't tell you that.
Me: HAL, how deep am I?
HAL: You're in deep ****, Dave. Would you like to hear a song? Daisy, daisy.....

 

[ianr33]

Does a dive shop make more money selling;
a) An spg?
b) An AI computer ?

How long before the dive industry starts pushing for laws requiring all divers to use an AI computer? ("surely it will be worth it even if it only saves one life..................." )

Maybe I'm just in a cynical mood today

 

[Hatul]

Does a dive shop make more money selling;
a) An spg?
b) An AI computer ?

How long before the dive industry starts pushing for laws requiring all divers to use an AI computer? ("surely it will be worth it even if it only saves one life..................." )

Maybe I'm just in a cynical mood today

I don't think there are many laws dealing with divers. You can do very stupid things and they're all legal.

As far as the AI computer debate goes, there is a bias against it among very experienced divers and I think it's because many are influenced by the tech diving community which does not use AI. Then there are the recreational divers who use AI and find it very useful and that it helps with safety vis-a-vis gas management. The other issue is that AI raises the price of the dive computer which can be a burden to some people.

A compromise is to start with a wrist computer which is AI capable and use an analog SPG. Then you have the option of buying the transmitter at a later time, and the SPG becomes the backup.

 

[Peter_C]

I have clients that bring their Air Z Nitrox & Smart Z dive computers in for their Free battery service.

I explain to them that their 10 year old dive computers are still good.

They do have an option of getting a Galileo Sol or Luna at 40% OFF if they are the original owner.

Can you do say that about a PC ?

Aren't you selling all kinds of Scubapro stuff for 40% off? Why would one need to be an original owner if you can just buy one for 40% off?

My Smart Z had to be sent in for a free battery change. I have also replaced the transmitter battery. It isn't ten years old yet, but I expect it to last longer than that or have some kind of replacement available at a substantial discount. Although I would probably put the money towards a Liquid Vision for their superior OLED technology. Of course a Shearwater Predator would be awesome too. No need for an AI, as one still has to calculate Min Gas since these AI's don't do it for you.

 

[-hh]

Let's see what I've gleaned from this 11+ page thread..

I would guess that the failure rate of the transmitter has dropped substantially as the technology has developed. I would guess that the first SPGs were not very dependable either. Probably as much discussion between divers using the SPG when it first came out and the ones that didn't like it because they had a J valve and the SPG could fail.

Correct on both counts: when first introduced, neither was particularly reliable in comparison to the 'legacy' solution.

This question is for single tank recreational dives only. Does anybody have 2 SPGs in case one fails? I don't see a huge problem here, you don't even have to have any special training to ascend without an SPG

A fair observation in that the different devices may very well result in having different 'recovery' modes that the diver would need to employ.

I wouldn't go on a once in a lifetime dive trip without having first put new battery in the computer and the transmitter and making sure it worked. Changing a battery prior to a trip would be cheap insurance and part of proper maintenance IMO.

Hope that your SOP includes pressure-testing that user-replacable battery, since a water leak here is probably the most common failure mode for electronic dive equipment. Of course, SPGs don't have batteries, so this failure mode doesn't exist.

I would also think that an issue with syncing has greatly decreased since the technology was introduced...

It has, although it should also be noted that this failure mode only applies to the 'hoseless' AI's. Of course, the 'hosed' AIs have the HP hose itself as a potential failure point ... and these may also include the same o-ring swivel gland that's a common leakpoint on SPGs, so for that application, there's not an additional failure point in the one versus the other when one tries to develop a failure model for each.

Unfortunately unless someone has data to prove one way or the other it is strictly personal opinion.

Data, which would include the relevant baselines and other 'denominator' types of factors. It is possible to build some design models to estimate how many major failure points there are, from which one could postulate some relative reliability factors.

For example, a hoseless AI has twice as many batteries (and battery compartments subject to leaking) as a hosed AI, so if that was the only reliability factor, we would expect that a hoseless AI should be twice as likely to have a failure as a hosed AI. Of course, this isn't the only design difference, nor the only failure modes, so this baseline estimate isn't correct.

It really would be interesting to have hard failure rate info on AI vs non-AI computers. However, I doubt that the scuba equipment manufacturers would want that knowledge published. The best we have is personal observation. I've seen enough for me. Not that I think having an AI computer is "wrong," they are convenient and have lots of whistles and bells. However, I'd want a backup computer and SPG.

Yes, it does come down to "who has the data?". Expanding briefly on my above comment of building a failure analysis model, we can come up with:

Hosed AI:
- HP O-ring @ tank
- HP hose
- HP swivel O-ring @ gage
- HP pressure transducer
- water pressure transducer
- battery compartment seal
- battery
- electronics package (high level)
- display electronics

Hoseless AI:
- HP O-ring @ tank (transmitter)
- HP pressure transducer
- battery compartment seal (transmitter)
- battery (transmitter)
- electronics package (high level)
- DC-analog converter & transmitter electronics
- water pressure transducer (receiver)
- battery compartment seal (reciever)
- battery (receiver)
- DC-analog converter & receiver electronics
- electronics package (high level)
- display electronics

Thus, the model's basic comparison is:
Hosed AI (9 things which could go wrong) vs. Hoseless AI (12 things which could go wrong)

...plus we could to build out a similar model for the dumb old analog SPG too. Then all we would have to do is to calculate the reliability values for each possible major failure mode and do some math. Of course, we don't have good numbers to work with - - but just for sake of illustrating the math, if we were to assume that all of these are the same, then the math simplifies (vastly simplifies!) down to:

Hosed AI: P(fail) = 1 - R^9
Hoseless AI: P(fail) = 1 - R^12

For a notional R (Reliability) value of 0.999, these would calculate to Probability of Failures of: 0.896% and 1.193% ... not necessarily a huge number per se, but it is worth noting that the one value is 33% larger than the other, simply because there's more opportunities for things to go wrong on the one design.

I have a question for you: Have you sold equal numbers of both?

I ask because if there are 1000 SPG's out there, and 250 transmitters, then even if you have twice as many SPG's failing as transmitters, that would still be a higher failure rate for transmitters.

Without knowing the total numbers we can't tell, but I suspect there are (at this point) many, many more SPG's out there, so I don't see "more coming in with problems" necessarily being meaningful.

There's more variables than just that too. For example, to provoke (humorously!) some stereotypes, the comparison also gets biased if the SPGs are coming from local greybeards who dive every weekend, whereas the AI's only get 10 dives/year from warmwater holiday vacation divers, this also influences the "denominator".

Another good question is one of what are the repairs for? If 90% of them are for leaking O-rings at the swivel, that's both a quick & cheap repair as well as more of a "nuisance" element - - it probably didn't cause a major failure that would have caused a dive to be aborted. This basically doesn't treat all repair requests as being utterly equal in significance, but looks at different modalities.

I understand how an AI computer calculated gas requirement to ascend on a standard profile, but could you please educate me on how the SOL has any idea on how much gas it takes to make his swim back (to the anchor line, out of the kelp, etc...)

To play Devil's Advocate, how did the same SOL know that it was a "5 minute swim" to be able to apply the data his AI gage was giving him?

We have been using UWATEC Hoseless Air- Integrated Dive Computers ( HAIDC ) in our training for both students, DM & Instructors since the late 90's. We have sold HAIDC's 10 to 1 ( SPG ) for the last 15 years.

To play Devil's Advocate (again), how many of these HAIDC's were "Repeat Sales" because their first HAIDC had died?

We are comparing apples to pizza. A hockey puck SPG is as simple as it gets, yet these things fail. An AI wireless computer is about as complex as they come offering up dozens of functions and configurable options. It's hardly a valid comparison.

It is a hard comparison to make good. Sure, an SPG is pretty dirt simple and with very few things to break, but it is also basically 100% mechanical, whereas the AI is a hybrid of electronics + mechanicals, and since E's are generally more reliable today than M's, when you reduce the number of M failure points, you can add a lot of E's without the overall reliability of the device getting worse. The net effect is that they're simply a lot harder to compare...and that's even before we try to account for their differences in failure modes, which do not necessarily all result in the same outcome: an HP hose's swivel's leaking O-ring isn't always an immediate stopper, as would be a full flood of the electronics package in an AI dive computer.

For whatever reason divers seem to want to use old technology. The rational seems rather flawed. In an age of computerized brakes on cars, and computer controlled everything many divers reject technology and I honestly can not explain it. Your life is literally dependent on technology on a daily basis yet many reject the very things the they depend on daily. I dive the Epic and have had unbelievably good luck with this computer. Granted I carry a backup. Everything electronic you subject to an underwater environment will break eventually. Your mileage may vary.

Speaking of mileage, some scientific studies have found that the introduction of Anti-lock brakes to automobiles as a safety device did not result in a permanent improvement in highway safety: it turns out that their improvement was offset through a Risk Compensation behavior of the operator. Diver aids are subject to this same risk.

Based on observations while on many dive boat trips, divers have problems with AI computers much more often than they do with simple SPGs. I suspect that the vast majority of "failures" aren't really failures of the transmitters, but are instead simply a dead battery.

That still sounds like an obstruction to making a dive, even though the correction was relatively inexpensive and remedial.

As far as the AI computer debate goes, there is a bias against it among very experienced divers and I think it's because many are influenced by the tech diving community which does not use AI.

That's one opinon. Frankly, I looked at them with my Engineering assessment hat on, years before the WKPP even came into being.

Then there are the recreational divers who use AI and find it very useful and that it helps with safety vis-a-vis gas management. The other issue is that AI raises the price of the dive computer which can be a burden to some people.

IMO, both are valid observations. Perhaps a better combination would be to say that those experienced oldtimers are a bit cynical of spending the extra money for AI when it doesn't do much to make their dives any longer, particularly since they've seen the dive industry go through plenty of fads over the past few decades.....which reminds me: has anyone seen my bright green wetsuit from the 1980s? I think I left it with my Force Fins and nitrox-cleaned Spare Air

A compromise is to start with a wrist computer which is AI capable and use an analog SPG. Then you have the option of buying the transmitter at a later time, and the SPG becomes the backup.

And YA compromise is to skip the AI portion and just use a non-AI dive computer. That combination is even cheaper, since an AI without a sender unit won't give any air consumption based "time remaining" guidance anyway.


-hh