Shearwater Perdix AI

[CarcharodonCarcharias]

I was hoping for a Perdix with AI, but couldn't wait and ended up buying a used VT4.1 less than a year ago. It works fine and does what I need it to do, so I'm going to hold off. This does look like the perfect DC, though... If I had know, I would have waited. Oh well.

Did you get the TX? If you did, just use the VT as the back up and buy the Perdix AI.

 

[scubadada]

I was hoping for a Perdix with AI, but couldn't wait and ended up buying a used VT4.1 less than a year ago. It works fine and does what I need it to do, so I'm going to hold off. This does look like the perfect DC, though... If I had know, I would have waited. Oh well.

Me too. I'm diving a 6 year old VT3, I got my son a minimally used VT4.1 about a year an a half ago, both are just fine for now. And.....I have 2 PPS transmitters

 

[the_ocean]

The VT4.1 is great, until it bugs out on you. On a nitrox dive series, it forgets to offgas during your surface interval if you use the plan or set FO2 feature between dives. Oceanic has replaced my computer twice with no root cause explanation, for this firmware issue that they claim to have fixed years ago. Details here:

Hollis TX-1 review

I like the VT4.1 other than this bug. I plan to pick up a Perdix AI and switch the VT4 to backup.

 

[stuartv]

Hi Stuartv,

This is not quite correct. With our implementation, you can see pressure from either transmitter at any time. Either transmitter can be placed on the main screen, or two right button presses (one step past the compass) brings up an AI line, that doesn't time out back to the main screen. You can see a bit of this in a video we posted.

Not linking the transmitter to a specific gas results in a simpler system that is more flexible. The downside is GTR being limited to one gas. Although we find 1 gas GTR is ideal anyways. We tested more complex GTR with multi-tanks, and the setup required to cover different scenarios properly is too high. Our test engineers made frequent mistakes or forgot to change settings, which resulted in incorrect GTR. I hope to cover the rational for our design decisions in a future video.

We appreciate all the interest, discussions, and ideas. Many of us are at DEMA now, but we will provide more information once back in the office.

Best regards,
Tyler Coen
Shearwater Research

Hi Tyler,

I really appreciate you taking the time to respond to my earlier comments.

I have to take a bit of issue with your statement that my earlier comment was not quite correct. I believe my statement was correct and what you've posted is that there is a workaround available, if you are willing to give up more of your screen real estate in order to have both transmitters' pressures displayed at the same time.

My earlier point was that, in the normal configuration that I would want to use (and am very accustomed to using with my current AI computers) there is only one field on the normal dive screen that is used for tank pressure. I like it that way. I am only concerned during a dive to see the pressure of the tank I am breathing. And I want to use the rest of the screen real estate for other information.

With the way you have implemented AI, if I configure my normal dive main screen to only show 1 pressure, then if I switch tanks (e.g. from 1 sidemount tank to the other), I would also have to change the configuration on the Perdix from showing one tank's pressure to showing the other.

A better example might be if I'm switching between a travel gas and back gas. I only want one pressure shown on my screen - the one I'm breathing. When I do the gas switch to/from travel gas from/to back gas, I want the pressure that is being displayed to change automatically to the new tank I'm breathing.

Ditto, if I'm switching from bottom gas to deco gas.

As it is, your current method is only even usable while you only support 2 transmitters. If you expand to allowing a transmitter on every gas, your workaround solution will not scale. I definitely would LIKE a screen I can go to that shows me all my gases, each with their tank pressure. But, for my main dive screen, I just want to see exactly one tank pressure - the one I'm breathing from.

That is how it works on my Oceanic and Hollis AI computers and I can't imagine a reason that I would not want it to always work that way.

Show me one tank pressure on the main screen and make it always be the gas I'm breathing (assuming i have a transmitter assigned to that gas, otherwise "----"). Simple. Intuitive.

For diving with only two cylinders (or even 3 or 6), Hollis and Oceanic have definitely made it simple enough to configure the gases and transmitters so that I have never had a problem with it. It's just not that hard.

Lastly, obviously just speaking for myself on ALL of this, I really don't care about having RMV (cu-ft/min) displayed on the computer. I only ever look at that after I download my dive data into a desktop computer log. So, at least as far as I am concerned, you could get rid of ALL the stuff in the Perdix related to configuring tank size. I can enter tank size data in the desktop logging software, so that the software can calculate my RMV.

If it's the combination of configuring tank sizes and assigning transmitters to specific gases that makes it too complicated, then I would definitely choose to drop tank size rather than dropping gas-transmitter assignments. Having my displayed pressure automatically update to the correct tank when I do a gas switch is a lot more useful to me than seeing my RMV during a dive.

I don't think there is any possible way of doing it that is simpler than "there is one pressure displayed on the screen and it is the pressure of whatever gas you are breathing."

As @scubadada has already mentioned, Oceanic and Hollis have been doing AI for a while. It works well. There is nothing wrong with looking at how they do it and doing it at least as well. I have a Hollis TX-1. It is not a great computer. But, the gas and AI management that it gives is, in my personal opinion, nicer gas management to use than my Petrel 2 had and way nicer AI management than what you've incorporated into the Perdix AI.

Similarly, it may look "funny" to your engineers to have a GTR of 0 while hanging at a safety stop... but they'll get over it. It makes perfect sense that if you stay on the bottom until GTR hits 0, that it would stay at 0 all the way to the surface, as long as you follow the ascent that the computer gives you. So, my desire would be for GTR to reflect any and all stops (mandatory or optional) that the computer is going to give me as I ascend. If the computer is going to tell me to do a 3:00 safety stop at 15', then the GTR it shows me at the bottom should reflect that I'm going to stop at 15' for 3:00.

If the computer is accurate and precise and I follow it exactly, doing every stop it calls for, that should mean that if I leave the bottom exactly when GTR hits 0, and I ascend at precisely 33'/min, do every stop at the precise depth, for the precise time called for, including the safety stop, and my breathing stays consistent throughout it all, then I should arrive at the surface with precisely the gas reserve that was configured in the System settings.

That seems pretty simple and the most intuitive way for GTR to work. If I do all that correctly and I arrive at the surface with a different amount of gas than my configured reserve, I would feel like my computer was not calculating GTR correctly.

And I certainly don't see any reason that GTR shouldn't be shown - and be correct (per preceding paragraphs) - even if I'm doing a deco dive with deco on back gas only.

I am thrilled that you have come out with the Perdix AI and that it works with PPS transmitters. I will be buying one, even if the AI implementation stays exactly as it is now. I'm just trying to communicate how I, as one customer, would feel like it would work even better for me.

Thank you again for being here and listening to all our input!

 

[scubadada]

The VT4.1 is great, until it bugs out on you. On a nitrox dive series, it forgets to offgas during your surface interval if you use the plan or set FO2 feature between dives. Oceanic has replaced my computer twice with no root cause explanation, for this firmware issue that they claim to have fixed years ago. Details here:

Hollis TX-1 review

I like the VT4.1 other than this bug. I plan to pick up a Perdix AI and switch the VT4 to backup.

I've never had that problem with my VT3 in over 6 years, over 800 dives. I use the planning function very frequently and have to change FO2 reasonably often. My VT3 will also eventually be relegated to backup, running off the same transmitter as a Perdix AI

Maybe it's a problem that occured later in the series, VT4, 4.1. My son has not had that problem with his VT4.1 but he has only about 20 dives on it. The Veo3 is a similar computer but without AI, the TX-1 is related but with many differences from the Oceanic computers

 

[Wah]

Did you get the TX? If you did, just use the VT as the back up and buy the Perdix AI.

Funny.... I was primarily using an Atom 2.0 w/tx that went to backup duty when I bought the VT4.1without a tx... so the backup role is already taken. I need a better excuse to dump my VT4!

 

[scubadada]

Funny.... I was primarily using an Atom 2.0 w/tx that went to backup duty when I bought the VT4.1without a tx... so the backup role is already taken. I need a better excuse to dump my VT4!

Maybe you "need" to dive Buhlmann with GF someday (same deco algorithm as buddies, availability of planning software, need for flexible, adjustable conservatism...) Just trying to help. You really only need to convince yourself, well, and maybe a spouse

 

[Hatul]

Hi Tyler,...

Similarly, it may look "funny" to your engineers to have a GTR of 0 while hanging at a safety stop... but they'll get over it. It makes perfect sense that if you stay on the bottom until GTR hits 0, that it would stay at 0 all the way to the surface, as long as you follow the ascent that the computer gives you. So, my desire would be for GTR to reflect any and all stops (mandatory or optional) that the computer is going to give me as I ascend. If the computer is going to tell me to do a 3:00 safety stop at 15', then the GTR it shows me at the bottom should reflect that I'm going to stop at 15' for 3:00.

If the computer is accurate and precise and I follow it exactly, doing every stop it calls for, that should mean that if I leave the bottom exactly when GTR hits 0, and I ascend at precisely 33'/min, do every stop at the precise depth, for the precise time called for, including the safety stop, and my breathing stays consistent throughout it all, then I should arrive at the surface with precisely the gas reserve that was configured in the System settings.

That seems pretty simple and the most intuitive way for GTR to work. If I do all that correctly and I arrive at the surface with a different amount of gas than my configured reserve, I would feel like my computer was not calculating GTR correctly.

And I certainly don't see any reason that GTR shouldn't be shown - and be correct (per preceding paragraphs) - even if I'm doing a deco dive with deco on back gas only.

I am thrilled that you have come out with the Perdix AI and that it works with PPS transmitters. I will be buying one, even if the AI implementation stays exactly as it is now. I'm just trying to communicate how I, as one customer, would feel like it would work even better for me.

Thank you again for being here and listening to all our input!

I can't comment on multiple cylinder diving as I only dive with one +- a pony. But I do hope that Shearwater includes the mandatory/safety stops in the GTR algorithm. As Stuart pointed out then the definition of GTR is very simple, for as long as the diver keeps the GTR above zero at any depth, breathing does not change, and he follows the prescribed ascent rate he should surface with the reserve pressure.

 

[Diving Dubai]

@stuartv

Why is everyone getting so bent out if shape over GTR? If it's that important, then perhaps SW should dispense with a gas contents display and just tell you how many minutes you have remaining.

Yes GTR is nice, and it should include all the stops - maybe for some that needs a TTS as well.

But in the real world, if people can't balance your gas, and figure out if they need to ascend to surface with reserve, or if they have enough gas to allow them to take deco, then why are they making those dives?

it appears this is a contradiction. People want GF to give them the decision making on algorithms yet want a computer to nanny them on their gas contents

 

[stuartv]

@stuartv

Why is everyone getting so bent out if shape over GTR? If it's that important, then perhaps SW should dispense with a gas contents display and just tell you how many minutes you have remaining.

Yes GTR is nice, and it should include all the stops - maybe for some that needs a TTS as well.

But in the real world, if people can't balance your gas, and figure out if they need to ascend to surface with reserve, or if they have enough gas to allow them to take deco, then why are they making those dives?

it appears this is a contradiction. People want GF to give them the decision making on algorithms yet want a computer to nanny them on their gas contents

I think it's as simple as people feeling like if it shows them a number, it should be as accurate as possible and available any time that it can be reasonably made available.

So, if it's going to have an option to show GTR, make it show a number that is as accurate as possible. And do the calculation in the way that is most intuitively clear to the user. I.e. in this case, if I stay until GTR = 0, then follow the computer exactly, on my ascent, I should arrive at the surface with precisely the configured reserve left in my tank. I would rather not even have a GTR display than to have one that doesn't reflect the stops that the computer is going to "suggest" that I make.

Do it right, or don't bother doing it at all.