So, Will Shearwater offer Gas Integration

Please register or login

Welcome to ScubaBoard, the world's largest scuba diving community. Registration is not required to read the forums, but we encourage you to join. Joining has its benefits and enables you to participate in the discussions.

Benefits of registering include

  • Ability to post and comment on topics and discussions.
  • A Free photo gallery to share your dive photos with the world.
  • You can make this box go away

Joining is quick and easy. Log in or Register now!

The engineers have already figured this out.

Yes, even now with current AI technology, the simple fact is that there is no documented evidence that AI increases failure of the deco functions of a computer. Just doesn't happen. If AI fails, the computer just keeps chugging along. If this were a "real" issue and not a mental exercise, the horror stories and lawsuits would have become legendary by now.

Also, AI "failure" is not an actual computer issue but refers to the loss of connection between the pressure sensor/transmitter and the computer. This means loss of the gas reading, nothing more. Most connection drops are temporary and the computer picks up the pressure reading after a short time (personally, I have never had the connection drop). This does not affect the deco calculations.

It might affect the remaining gas time calculations, but, again, not the deco calculations. However, if you are just using the computer as a more convenient pressure readout, then those gas calculations don't matter.

All of this assumes, of course, that you dive with a back-up SPG (which I do).
 
The engineers have already figured this out.

Yes, even now with current AI technology, the simple fact is that there is no documented evidence that AI increases failure of the deco functions of a computer. Just doesn't happen. If AI fails, the computer just keeps chugging along. If this were a "real" issue and not a mental exercise, the horror stories and lawsuits would have become legendary by now.

Also, AI "failure" is not an actual computer issue but refers to the loss of connection between the pressure sensor/transmitter and the computer. This means loss of the gas reading, nothing more. Most connection drops are temporary and the computer picks up the pressure reading after a short time (personally, I have never had the connection drop). This does not affect the deco calculations.

It might affect the remaining gas time calculations, but, again, not the deco calculations. However, if you are just using the computer as a more convenient pressure readout, then those gas calculations don't matter.

All of this assumes, of course, that you dive with a back-up SPG (which I do).

Bring out the torches and pitchforks :angrymob:

[video=youtube;zrzMhU_4m-g]https://www.youtube.com/watch?v=zrzMhU_4m-g[/video]
 
AI as mentioned in the divecan is an entirely different concept than what most people think of when AI. You saw Bruce himself say that wireless AI isn't in the plan. So C dub, guess you can try to fire the owner of that company. You don't understand the purpose of that company, it is not a corporation out to maximize the profits, it is a small niche manufacture set out to make the best possible dive electronics out there.

The AI that could be incorporated into the Divecan, would likely be a HP quick disconnect, that plugged into the head, into a digital pressure sensor. No different than the ones used in the QD AI computers. It would output to the handset or HUD as just another data point that it would read out. That would require nothing other than a software update on the controller and monitor, but they would have to be wired in, same as they are to monitor PO2.

Actually, I would like to understand more about the dive can system. The dive can website had no explanation at all of how it works and neither does the Shearwater site. Can you give more info, or perhaps a link to something? There must be some type of pressure sensor and transmitter at the tank, and a receiver at the monitoring end, but the tank units shown did not look like typical AI transmitters, so I was wondering what different tech was involved.

---------- Post added November 18th, 2014 at 12:42 PM ----------

I prefer "the comfy chair . . ."
 
The engineers have already figured this out.

I'm sure they THINK they have it figured out, and by the time a new product is released they THINK they have tested it extensively. But every system is different. Just because failure of the AI system in one manufacturer's product didn't affect the deco system in that product doesn't mean it cannot happen in another manufacturer's product. I think the likelihood of some software bug within the AI system affecting the operation of the deco system is very remote--it SEEMS straightforward enough to keep the two systems independent of each other--but nothing is impossible. Who knows what software components the two systems might share. The only way to guarantee failure of system A or its components cannot affect system B or its components is to omit system A.
 
The engineers have already figured this out.

Yes, even now with current AI technology, the simple fact is that there is no documented evidence that AI increases failure of the deco functions of a computer. Just doesn't happen.

Actually, that's simply not the case. As an engineer, let me just tell you that few engineers have anything figured out :D.

Jokes aside, AI failures cannot be fully separated from deco algo failures. A failure in one can and will, in fact, lead to a failure of the other.

As far as no documented evidence that AI increases failure of deco functions, this goes back to them not being separate. However, there are plenty of AI computers that fail regarding their deco algorithms. There are plenty of computers that have an AI-glitch and then lock up for the rest of the dive. There are plenty of AI computers that actually use AI information in their deco algorithm (higher SAC rates means higher on-gassing, etc). They're simply NOT two separate systems. If they were truly separate, they'd be two separate devices.

Any time you add any features to any software/firmware it adds failure points. Period. Most dive computers are not known for their insane robustness, anyway....adding Angry Birds for safety stops just adds failure points.
 
The engineers have already figured this out.

Yes, even now with current AI technology, the simple fact is that there is no documented evidence that AI increases failure of the deco functions of a computer. Just doesn't happen. If AI fails, the computer just keeps chugging along. If this were a "real" issue and not a mental exercise, the horror stories and lawsuits would have become legendary by now.

Also, AI "failure" is not an actual computer issue but refers to the loss of connection between the pressure sensor/transmitter and the computer. This means loss of the gas reading, nothing more. Most connection drops are temporary and the computer picks up the pressure reading after a short time (personally, I have never had the connection drop). This does not affect the deco calculations.

It might affect the remaining gas time calculations, but, again, not the deco calculations. However, if you are just using the computer as a more convenient pressure readout, then those gas calculations don't matter.

All of this assumes, of course, that you dive with a back-up SPG (which I do).

Great theory, and probably correct. However, under it one is forced to assume that the reason that the AI failed was due to either a loss of connection or a failure of the computer -- you can't know that for sure when part of your computer's function goes on the fritz during a dive, though. Personally, I'll avoid the issue even though it surely leads to better gas consumption logging.
 
I'm sure they THINK they have it figured out, and by the time a new product is released they THINK they have tested it extensively. But every system is different.

Yeah:
ebbmv.jpg
 
Tot a certain point it can be done. But the fact remains it is running on a single operating system on a single computer. So there always is a risc.

Verstuurd vanaf mijn D6503 met Tapatalk
 
Despite the comments above, no one has pointed to a single concrete examples of deco algorithm failure related to AI. This lack of a relationship between AI and deco algorithm failure seems to be a fact, not a theory. The only theory, that there "might" be a relationship between AI and deco algorithm failure, has not been supported by observation or evidence and so, under the scientific method, has not been established as a valid theory.

I have seen no evidence that (1) an AI glitch "locks up" a computer, or (2) that computers using AI info to adjust to a more conservative factor cease to function if the gas info is lost (in fact, my Galileo will continue to function, but I do not use the heart rate/gas consumption adjustment to the algorithm because I do not know exactly how it adjusts, or what the basis is). If this had actually happened and people got bent it would have been reported all over. I have not found anything on it, anywhere. If there is actual, tangible evidence on this then please let me know as I will always re-consider if a factual basis exists to do so.

With all of the evidence pointing to a conclusion that AI does not affect deco reliability, the concerns here must be deemed theoretical concerns that should not block implementation of this feature.

Finally, let's not lose sight of the purpose of my original post, which is that the good folks at Shearwater think that some type of AI technology is reliable enough to implement from a diver to a far distant monitoring unit, which will simultaneously be monitoring the diver's deco status, so I rest my case.
 
Despite the comments above, no one has pointed to a single concrete examples of deco algorithm failure related to AI. This lack of a relationship between AI and deco algorithm failure seems to be a fact, not a theory. The only theory, that there "might" be a relationship between AI and deco algorithm failure, has not been supported by observation or evidence and so, under the scientific method, has not been established as a valid theory.

You're confused about who has what burdens here. The only fact established in our scenario is that an AI computer has stopped functioning in the manner in which it is intended to function--we know nothing else for sure about the nature of the failure. The two basic possibilities are (A) the failure has compromised only the AI function, or (B) the failure has compromised more than just the AI function...you're assuming that because generally failures only compromise the AI function, that is always true. That may be a fairly safe assumption given how infrequently your experience shows failures of AI to mean failures of anything else like deco calculations, but it's still an assumption one need not make if there's no AI functionality to fail in the first place.
 
https://www.shearwater.com/products/swift/

Back
Top Bottom