Having created a dive proof device that can accurately measure and record magnetic bearing, depth and other useful information like temperature, pitch and roll I was keen to see how it might perform as used to plot the route covered underwater by a diver. It was still without distance measurement capability at this point but assuming a fixed speed is maintained then this isn’t an issue, so I chose to strap it to the front of my Aquazepp and do a test to see if this idea was worth further pursuit with the hardware I have.
A device like this has been created before and carries the nickname of the ‘Lazy Boy Mapper’ developed by John Volanthen around 10 years ago and was used to great effect in long and deep caves where normal survey is not practical . Commercially available devices are also available by the like of Seacraft though these stretch into the £1300 region.
http://www.pozoazul-cavediving.org/index.php?option=com_content&view=article&id=16&Itemid=58
Around Christmas 2020 we were emerging from tier 3 lock down but yet to be dropped into tier 4 so a flurry of quarry diving activity took place due to the weather and ban on entering Wales for fun purposes.
I reprogrammed the device to record data continuously which differs to how it was setup to work when used to record lines used for cave diving. Still lacking a form of distance measurement this was intended as a test to see how it would perform and if this idea was worth further pursuit, scootering was done continuously and in one speed so that a fixed estimate of speed could be applied later on.
A rather chilly two laps of Cromhall Quarry were done (in a semi drysuit) at maximum speed. The data is quite encouraging for a first attempt.
The dive begun with a lap around the perimeter before looping through the container and maintaining a fixed bearing across the middle section at a deeper depth, this was followed back on reciprocal where another loop through the same container was done, a reverse lap around the perimeter was completed before finishing with another loop through the same container before exiting.
The ability to scoot through a fixed object like a container provides a reference point to where the data should match up, without this it is difficult to know where you have covered the same ground which allows an assessment of accuracy to be made.
In the image below it can be seen that the start and the end of the dive don’t agree location wise even though i did start and finish the dive in the same place, the rest however appears to have worked very well ! A fixed speed of 60 meters per minute was applied. Speed through the water will vary slightly with subtle changes in diver position throughout the dive so differences are to be expected.
The magnetometer quality factor which should remain fixed on a value of ‘3’ after calibration was changing throughout the dive so this might have been caused by the proximity to the batteries or motor on the scooter so this might be a cause for the drifts seen at the start and end of the dive, it seems to cope with short exposure of large metal objects like scootering through metal containers quite well (it is an intelligent sensor not just a compass so should be able to filter out temporary disturbances in magnetic field).
Future testing and improvements will involve:
Mounting the box on the nose of the scooter, as physically far as possible from the diver, motor and batteries.
A form of speed/ distance measurement will be added.
Calibrating the IMU whilst the box is mounted to the scooter in the hope that it accounts for the field of the scooter.
Align the heading of the box better with the heading of the scooter.