Tuba - The Open-Source Glider

Description

Despite the admirable efforts made so far towards a low-cost open-source underwater glider, currently the only way to get a glider is buying a commercial unit which costs around 200k USD. My aim is to create a blueprint for a glider where students and hobbyists from all over the world can build (and help improve) in order to explore and study down to a 200m (and hopefully at a later stage 1000m) of water autonomously for weeks or even months at a time without using an expensive commercial glider. I have also started a wiki about everything gliders which is linked below.

Details

An underwater glider works on a similar principle as normal air glider where the medium is water. There is no propeller in either glider, but the main difference is that unlike air gliders, underwater gliders use a neat trick to climb back the water column after gliding down. They do this by changing their density. Gliders are carefully designed to have a density very close to that of water. Then by simply changing their volume (which can be achieved using a simple piston pump or an inflating oil bladder), they can change their density and use the buoyancy force to fly up and down the water column.

By moving a large weight internally (usually the battery pack) it can control its pitch and by rotating the weight, it can roll and therefore fly towards a certain heading.

The glider will consist of three separate sections. These are buoyancy drive, mass shifter and electronics sections. Each section will be housed inside a 5" BlueRobotics pressure housing. This way, later we can upgrade the buoancy drive section to reach the full 1000m depth.

Initially I will focus on making a simple but robust buoyancy drive system. This will inistially consist of a custom-made rolling diaphragm and a linear actoator. Next step is to focus on the mass shifter system. As for comms, I am planning to use GSM and LoRa at the beginning and add Iridium at a later stage. The aim of this first build is to quickly and cheaply put together something functional and see what major changes needs to be done for the next version.

Files

Tuba_step.zip

Full assembly in step

x-zip-compressed - 37.83 MB - 02/18/2026 at 21:15

Download

Tuba.f3d

Full Fusion assembly

fusion - 31.32 MB - 02/18/2026 at 21:14

Download

Project Logs

Collapse

Almost ready for the first test deployment
Ehsan Abdi • 02/18/2026 at 20:56 • 0 comments

Moving the more detailed week to week logs to Discord. But I will try to post some updates here as well from time to time.

I did some major changes to the mass shifter assembly

and made a proper battery pack for it.

Added limit switches for pitch as it's the most critical part that really needs them for now but eventually roll, pitch and the pump would all need to have pots to keep track of where they are.
Had a friend make fiberglass fairings and we almost went for a deployment, but then decided not to rush it and make some more refinements.
Updates on mass shifter and VBD
Ehsan Abdi • 08/11/2025 at 19:14 • 0 comments

For the mass shifter, steppers weren't strong enough so switch to geared DC motors. Since these motors are slightly larger, I needed to switch to the 50cm housing instead of the 30cm one.
It's a quick and dirty design but does the job for now.
For the buoyancy drive, I played quite a bit with different types of molds for the diaphragm and finally managed to make one in silicon that has no holes. As for the seal, I used this parametric spring tool to make a large snap ring that would push the diaphragm to the housing and create a seal.
This together with a little bit of silicon adhesive did the job and I managed to pull a partial vacuum to test for leaks. Then I tested it in water and it does work! Managed to test it down to 5 meters but only have this shot from the initial test.
Mass shifter first try
Ehsan Abdi • 02/17/2025 at 19:38 • 0 comments

I'm still making some slow progress. I'm still experimenting with the design of the mold for the rolling diaphragm seal. I haven't managed to get the result I am looking for just yet. As for the attachment of the seal to the housing, I am now thinking about using some type of snap ring to exert a constant pressure on the walls and create a seal like that. But since I'm waiting for a new order of silicone rubber to arrive up here in the Arctic, I started experimenting with the design of the mass shifter section. I thought it would be easy enough to simply use NEMA steppers for both roll and pitch. I found this cool herringbone planetary gear which would do well for roll mechanism.
After printing and putting things together and some testing, it seems llike the NEMA steppers won't have enough torque to do the job, at least not reliably. So I will probably move to geared DC motors and potentiometers.
Testing the first design
Ehsan Abdi • 08/27/2024 at 21:21 • 0 comments

I just printed a couple of parts to quickly try everything together.
The hardest part was to glue the seal to the housing. The seal is made of silicon rubber which does not really adhere to anything. But finally, using some silicon-base adhesive I managed to quickly test things together.
Making a Rolling Diaphragm Try#1
Ehsan Abdi • 07/11/2024 at 13:46 • 0 comments

The idea was to try to make a rolling diaphragm the easiest way I could think of and see what happens. Since it needs to be flexible, I decided to make the first out of silicone rubber. I used the housing itself as part of the mold and 3D printed a few more parts to act as the rest of the mold.
It didn't turn out too bad but I think I will try another way next time and not using the housing as a part of the mold because I had a really hard time removing the piece!