Software configuration for an ardupilot based underwater glider

Hi Everyone,

I am a university student with considerable arducopter experience looking to work on an underwater glider for my master’s thesis. My background being in mechanical engineering, I am unsure how to go about developing the firmware for such a project. I am looking for some similar, well-supported firmware which will limit the amount of coding I have to do and allow me to focus on the development of hardware. I have searched far and wide for any kind of open source framework which supports underwater gliders and cant seem to find anything promising. What I have in mind is an underwater glider with a buoyancy engine as well as some way of changing its course (rotation of the battery within the casing). In an ideal world, I would be able to integrate a sensor for detection of the sea bottom as well as a GPS and satellite link. This would allow the glider to navigate autonomously over large distances, as well as to send some sort of updates about its location back to us. (basically a fully-fledged glider). My goal is to limit custom software development to the absolutely necessary minimum. I am looking for any firmware, frameworks, and sensors/microcomputers developed for my application.

Thanks in advance!

Hi @hugokopter -
A DIY glider sounds like a cool project! Unfortunately, you’re ahead of the curve - no autopilot firmware that I’m aware of supports your application. Some significant development may be required! Some groups have had success with using the Navigator WebAssistant to develop swarm AUV controls using our existing Raspberry Pi 4 + Navigator.

I would also point out that your underwater vehicle won’t have GPS link, or the ability to talk to satellites, unless it is at the surface of the water!
Best of luck!

Hi @tony-white, thanks for your time. That is what I was worried about… I guess i will keep looking. About the comment on the gps: the gliders surface every few dives to establish their gps location and chart a heading for the next few. This is also when all of the satellite communication happens so no need to worry about water not letting the signal through.

Hi @hugokopter, welcome to the forum! :smiley:

A similar question was asked about recently on GitHub, and my response there may be worth referring to as a starting point if you’d like to go the ArduPilot route.

Alternative/advanced control mechanisms are certainly exciting to consider, although as @tony-white mentioned you’re sufficiently ahead of the curve here that the kind of vehicle you’ve described would likely require some significant development work to be feasible.

Hi @EliotBR, thanks for your reply. In this case i think i will have to settle on writing simple code to run it myself using python and ROS or something similar. Its interesting that something like this is not supported as gliders are extremely widely used platforms, just mainly in academia which is not a wide market itself. Thats why probably there is no incentive to develop firmware for them.

Hi @hugokopter -
While some folks have developed software that uses the Navigator to control AUVs, you’re correct in that most AUV / Glider applications have not shared this development, and it is a fairly limited market compared to surface and ROV vehicles…

hi @hugokopter, I’m in the process of developing a subsea glider as well. I was also intending to approach this problem with Ardupilot as well, but it seems I’ve come into the same problems as you. I’ve spoke to someone on the Ardupilot form who’s had some success using ardurover and a companion computer for the dives, but there doesn’t seem to be one perfect answer out there. let me know if you’d like to discuss some time, I’m curious to see how you’ve adapted to this problem as well.

Great to see more people are getting interested in gliders! I have also started a glider project recently which you can check out here: Open-Source Glider | Hackaday.io
I have been working with commercial underwater gliders for over a decade now and I really want to see them be accessible to anyone! They are absolutely amazing and a fantastic tool for long-term study of the oceans.
@hugokopter how is your project going so far?
Regarding the autopilot, you need to think about it differently than an ROV or even a propelled AUV. The main point of a glider is that, it’s extremely efficient and low-power! So milliwatts matter!
What commercial gliders currently do, is that there is a very low-power supervisor microcontroller which is in charge of the flight-related decisions (which can take up to 7 hours if you are diving down to 1000m for example) and then when glider surfaces, it wakes up the power-hungry cousin for taking care of the file transfer and whatever else needs to be done. Also only a few kbytes of data should be transferred to decrease costs (especially if you are gonna use Iridium eventually for open waters) and minimize the time at surface.
Great to see your message here @brad_smith ! We forgot to arrange a meeting after your message in the Glider community Slack channel. We definitely need to organize a call and add whoever is interested to talk about this project!

2 Likes