CaveDiveMapping app and device

[Andrey Manolov]

Hello forum,

I have been developing a cave survey app specificly for diving.

Basically it's a cheap and open source mnemo survey/mapping device.

App is free and measuring wheel is 3d printed. Diving case is cheap 40 eur.

The solution uses an iphone's internal sensors to survey a cave with reasonable accuracy.

Goals:
1. Have an affordable survey gathering device for people to learn to map caves
2. Have entirely 3d printable hardware, so you can print it anywhere without needing wierd springs, bearrings or bolts
3. Have it gather as much data as possible only levaraging the iphone's sensors

Required parts:
I use PULUZ dive case for iphone 15 from Aliexpress. I see no reason why any dive case would not work.

The IOS app is free and open source as well as the STL and fusion 360 designs.

App source is avialable in github at:

GitHub - f0xdude/CaveDiveMap: An IOS app for taking survey data of underwater caves.​

An IOS app for taking survey data of underwater caves. - f0xdude/CaveDiveMap

github.com
(I don't have apple developer account so I can't publish in the app store)

3d printed device:

Cave Mapper Tool V3 by andrewxbg​

This is an updated version of the cave mapper tool for the CaveDiveMap App cave diving Survey tool.This tool uses a measurement wheel with a small magnet to measure the length of a cave diving line.An app on the phone in a diving case detects each revolution by the magnetic field and saves the...
www.thingiverse.com

Whoever is interested in mcguivering ****, please feel free to pull/fork the code and test it.

The app is available in App store:

CaveDiveMap App - App Store

Download CaveDiveMap by ANDREY RUMENOV MANOLOV on the App Store. See screenshots, ratings and reviews, user tips, and more games like CaveDiveMap.

apps.apple.com

Since, in Bulgaria we have very small number of cave divers real world testing is slow. I invite anyone who is intersted in the subject to print the wheel, and download the app.

 

[Chris Curry]

Looks great, well done Andrey! Does the magnet in the wheel not interfere with the compass readings?

 

[Andrey Manolov]

hey thanks, yes it does. The work around it that the app is constantly calibrating the compass. so when you reach a tieoff you rotate the phone around until the accuracy marker loght is green. the. you align the phone with the line and save manual data point. this way you have accurate reading. then you start measuring the line section where you dont care about interference. I will post a short video soon about the actual survey method as a tutorial.

 

[Andrey Manolov]

Here is my first map using the app/device.

 

[Andrey Manolov]

There is also an experimental feature for fully automated Visual Inertial Odometry and mapping which creates a point cloud in 3d. you can test this even in the dry cave. idea is using only visual processing as lidar is absorbed underwater.

 

[Andrey Manolov]

 

[Jim Warny]

Very interesting.

I have built a similar device with an arduino, and it uses a wheel with an IR reciever and emitter as the encoder.

Maybe you could do something similar using the phone camera to measure the turns of the wheel

How do you measure the depth?

Regards

Jim

 

[Andrey Manolov]

Ah cool. I had similar project 10ish years ago with arduino. I used an optical flow sensor from a computer mouse it sorta worked but never made it to the water.

I tried using the camera and also acoustic detection using the microphone each time the wheel turned it makes a clicking sound.
They both were giving me false positives.

Its definitely possible but i was not able to dial in the algorithm well. So for now i stick with the magnet wheel. IOS constantly calibrates the compass so when i reach a tieoff point i rotate the phone a bit when the wheel is starionary before taking the next heading.

I am also working on a visual mapping based on Visual Inertial Odometry SLAM. Its promising but still very experimental.

Give it a go it works for dry caves for now. No lidar only visual processing and accelerometer fusion.

 

[Andrey Manolov]

Sorry to your second question depth is entered manually. you look at your computer and select the depth with + and - buttons in the app at each station/tieoff

 

[Andrey Manolov]

Here is my buddy tring it for the first time yesterday. Works quite well. My idea was to not use any external hardware or as little as possible.