Wrist Mounted Computer

[rongoodman]

It's just market speak.

 

[scubadada]

Are there any references available to read to understand what the terms "folded" and "fully iterative" mean in this context please?...

I'd also like to read more about the Atomic algorithm. I seem to recall a post by Ron that attempted to explain some of the features to us, but can't find in it.

Ron, is there something we could read about it?

Thanks, Craig

 

[jd950]

Yep, that’s all pretty accurate. March/ April 2014 was when Liquivision started seriously being integrated into the Huish group. We actually had a Cobalt wrist version ready to prototype and test at the time, but put it on hold to work out how we were going to change to adopt some of the new technology. Turns out there was a lot we wanted to change to make a dramatically better product. There were about 10 months of business strategy related (not technical) delays in there (see November 2014), integrating all this growth was not easy. But for the last 16 months we have been working full bore on this revised version. Believe me, no one is more frustrated about the time this has taken than I am. In this day of apps, we tend to think in terms of short development cycles and knocking things out on a laptop, but this is bare metal engineering, hardware and firmware, for a dedicated real time system. It’s far more difficult and time consuming than writing applications, and requires a lot more testing. We’re testing now, and the project is very much alive.

-Ron

Glad to hear that. I look forward to seeing a product.

 

[RonR]

I’m really busy right now, and will try to put something more extensive together later. But in an over simplified nutshell:

Bubble models, of which RGBM is one, use known physics to model bubble formation in the body, attempting to formulate an ascent schedule that prevents bubble seeds from developing into actual bubbles that would need to be re-absorbed. The calculations are iterative and very computationally intensive, and until recently devices like dive computers just didn’t have the horsepower to do them. Even with today’s much more powerful processors, fully iterative bubble algorithms take some very efficient coding and firmware architecture if they are to give results in real time... We could have a longer discussion here about what constitutes a “real time system”- some examples being critical instrumentation on aircraft, machine controllers, or dive computers- as opposed to something computationally powerful, like a smart phone, that isn’t designed to operate as a real time system. But onward to answer your questions.

The “folded” RGBM algorithm uses Haldane / Buhlmann type calculations (using M-values)- these are much simpler. It then overlays bubble factors derived from running the “full” algorithm on powerful computers. This approximates the results of the full algorithm, while requiring much less power from the processor and being a lot easier to code. All recreational “RGBM” computers that are or have been on the market (apart from the Cobalt) to my knowledge are using a folded algorithm. Fully iterative RGBM (and VPM, another bubble model) software is available for some technically oriented computers as an aftermarket purchase, but (again, to my knowledge) none are shipping with it.

The Cobalt uses both. For dives shallower than 150’ we use a folded algorithm, which is set to be middle of the road as to conservatism (and is user adjustable within boundaries). Once you pass 150’ the Cobalt kicks in the fully iterative RGBM calculations- advantages using fully iterative calculations for shallower dives would be nonexistent, but we do have this ability for deeper profiles. This does yield better (more accurate per the algorithm, not necessarily shorter) schedules on such dives.

So, bottom line, they are different computational methods using different levels of approximation. One is much more challenging to implement. Hope this makes sense as a very truncated explanation.

-Ron

 

[medic29]

Just inquiring about quarterly update?

 

[RonR]

We're still working on details, particularly related to the appearance of the case. I can tell you I dove one yesterday, but there is no ETA for the released product, and probably won't be until Atomic is essentially ready to ship.

Ron

 

[MikadoWu]

We're still working on details, particularly related to the appearance of the case. I can tell you I dove one yesterday, but there is no ETA for the released product, and probably won't be until Atomic is essentially ready to ship.

Ron

Here is to hoping you all add tri-mix tot he computer. If my cobalt 2 had it, I would have keep it.

 

[scubadada]

We're still working on details, particularly related to the appearance of the case. I can tell you I dove one yesterday, but there is no ETA for the released product, and probably won't be until Atomic is essentially ready to ship.

Ron

That's a lot better than Liquivision has done with the Omnix, looks like a year from announcement to availability

 

[RonR]

Here is to hoping you all add tri-mix tot he computer. If my cobalt 2 had it, I would have keep it.

Atomic thought there would be little interest in tri-mix for a console computer. But the underlying algorithm is tri-mix, there just isn't a user interface for changing He% on the Cobalt 2.

Ron

 

[MikadoWu]

Atomic thought there would be little interest in tri-mix for a console computer. But the underlying algorithm is tri-mix, there just isn't a user interface for changing He% on the Cobalt 2.

Ron

Any chance then of adding the feature to cobalt with a future upgrade?