Underwater GPS

[stiebs]

For a patent, having it lodged and pending is sufficient to protect your IP. To me, it smells fishy that they have got it working so accurately (I assume the quote from the video of it being only 2% accurate was a slip but have not offered it up for independent review, or even a more in depth video or document of how it is employed.

Their website reeks of shallow marketing guff. They are apparently at BOOT, so maybe any SBers that are going can suss them out in person.

If I compare another location aware device (intelligent handheld CNC machine), albeit with a relatively simpler task and in a different field (woodworking), the Shaper Origin team have approached their market in an entirely more believable approach. Right from some of the early prototypes, they were sending their device out to popular woodwork youtubers, getting their name out there, but also proving through independent review that their idea and prototype did what it promised.

Ariadna could do much more than offer lip service, especially around a field that is so complicated and challenging to solve.

In the meantime, I'm going to a Google image search for ariadna and take some time to myself.

 

[NAM001]

@KWS, yes , I realise my accelerometer phone examples are in a different league to navigation, which actually supports my position from previous posts. The point of that post was in response to your comment that phones don't have accelerometers. They absolutely do, but they are not used for any sort of navigation because it is simply not viable.

A smartphone uses a combination of GPS, WiFi and 3G/4G triangulation for location services. It will always use GPS unless unavailable when it will fall back to the other two.

GPS is also used for elevation.

Accelerometer is used for the things I mentioned above, plus other minor movement detection such as turning the screen on if you pick it up (more common example in the Apple Watch, which turns the screen on when you turn your wrist)

I'm in absolute agreeance that in to to achieve what these people are trying to do requires a combination of many types of sensor technologies, and will be absolutely blown away if their current prototype can actually do it, let alone a miniaturised version.

A dead giveaway for me so far, is that they sell the wonders of it, and say it is working, but so far I have not found any evidence of them offering it up for independent use or review. If they are that confident, it should absolutely be patent protected and available for external review / input

My comment was based on you dont have accelerometers in a cellphone for navigation (since that was the general topic). Most likely with out them your screen would not change orientation, but that is not a nav issue.
The multi source of wi fi ect for positioning is a mixed bag of outputs. gps is of the accuracy needed to navigate. if gps is off you can get a rough position using the alternate methods you listed.
Their device no matter how it will work is constrained by physics. a larger device with greater seperation between motion sensors will be more sensitive to motion, hence greater accuracy. To achieve the same output from a smaller format gets expensive. Set a price of 2k for a unit and let that guide the technology you can implement on the device. After all you can compare the forces on 2 sensors X and Y and determine an angle of force direction. The question is can you make devices that are that matched to apply a sin or cosine function to at a reasonable price. You have to be accurate or the other option is to not have the accuracy on a sensing unit and implement several units that between them via an algorithym reduce the percent of error to a working level. 11 sensors were mentiioned. 11 cheap sensors at 100.00 can do what a couple of great sensors at 1000 dollars can do. slop is acceptable because of the non critical use of the device. I would also expect that in the implementation of this device it woudl be expected to get you to a reference point where you can take over. IE it can get you with in 20 ft from a desired location (big rock) and you know that 5 ft from there in the direction of the little rock is the burried treasure. More than accurate enought o find an anchor line or relative area of a planted tide level sensor of some sorts.

 

[NAM001]

I was trying to determine a cheap way to integrate a compass function to provide initial base posisiton and direciton. I could not do it but am assuming that a predive startup woud do that. you have gps you have location and if you move while holding the divice as you wold use it you get direction. Press a set button and it can run on accelerometers form that point as teh devices react to changes so i can tel you to go straignt for 100 miles turn right for 50 miles and then left for 75 miles, you can figure where you are if you know the starting position and the initial direction you are going to apply the changes to. such a benefit to using a compass that is effected by to many things in nature.

 

[stiebs]

A compass device already exists, and is used in many dive computers already. But a compass, especially such a small form factor with unknown and variable magnetic influences in such close proximity is all but useless in this context.

 

[NAM001]

A compass device already exists, and is used in many dive computers already. But a compass, especially such a small form factor with unknown and variable magnetic influences in such close proximity is all but useless in this context.

I agree completely, that is why i said that with an initial reference location and direction of movement you can eliminate an actual compass because using the motion sensors you can derive a compass function that are not a function of iron deposites and ship hulls. The compass in my shearwater depends on this requires a formal calibration to make it work. I wonder what would happen if I did the initial cal pointing other than north, would it then make IE east its calibrated north. ( i havnt done this) I MAY TRY IT. Either way if yo are going to power the sensors and you can get the same results that you would et with an additional compass inteegrated then why put the compas in it if you can calculate it.

 

[Captain Swoop]

Compass is easy, I have a Casio watch with a very accurate compass built in.

I would like to know how the device copes with being flapped around on someones leg.

 

[tursiops]

I believe the Office of Naval Research built a hand-held diver-tracking system for the SEALs, many years ago. It used accelerometers, gyros, and a small low-power high-frequency sonar device that looked at back-scatter from either the bottom or particles in the water. The position came from a Kalman filter based on double-integrating the linear and rotational sensors, with the sonar giving speed over bottom/through the water. It, I think, worked pretty well, but because it had batteries and o-rings, the SEALs did not trust it. If built today, it would be much smaller and cheaper. I admire the spunk of the Ariadna folks, but question their methodology.

 

[jseyfert3]

To me, it smells fishy that they have got it working so accurately (I assume the quote from the video of it being only 2% accurate was a slip but have not offered it up for independent review, or even a more in depth video or document of how it is employed.

What do you mean? 2% is what they claim. In the FAQ it says the same, 10 m off in a 500 m long swim, which is 2%.

In the meantime, I'm going to a Google image search for ariadna and take some time to myself.

 

[stiebs]

What do you mean? 2% is what they claim. In the FAQ it says the same, 10 m off in a 500 m long swim, which is 2%.

They quoted an "accuracy" of 2%, not an error margin of 2%.

 

[lowviz]

I admire the spunk of the Ariadna folks, but question their methodology.

Me too. It really looks like another classic "software solution to a fundamental hardware problem". We can either wish them failure or attempt to help.

So let's look at the fundamentally established approach of PID control. "P" is proportional, how far off am I? "I" is integral, how long have I been off? "D" is differential, how fast are things getting away from me?

Ariadna is lacking "P". How far off of the reference am I? Dead at the starting gate.

The reference:
1) On a drift dive, Ariadna is a slam-dunk. It knows where you are with respect to just drifting. Piece of cake.

2) On an Earth-based (fixed point on the bottom) referenced dive, you are screwed. Flying blind with no hope of correction.

I believe the Office of Naval Research built a hand-held diver-tracking system for the SEALs, many years ago. It used accelerometers, gyros, and a small low-power high-frequency sonar device that looked at back-scatter from either the bottom or particles in the water. The position came from a Kalman filter based on double-integrating the linear and rotational sensors, with the sonar giving speed over bottom....

Sonar/LIDAR doppler will give you a reading of your velocity with respect to the bottom, and thus, your point of entry/exit. Either cheap Sonar or LIDAR can be had for less than $200 US.

You need this and you know it...