Being an avid scuba diver, I’ve wanted to build a sonar to find things to dive on, as well as to document local dive sites, wrecks and reefs. Although side-scan would be the quicker and easier way to achieve this goal, I have set my sights on a Synthetic Aperture Sonar (SAS). The trick? I need to have some sort of stable platform to mount the sonar — And hence my Autonomous Underwater Vehicle (AUV) project. (In case anyone’s curious, I don’t have a boat, and boats aren’t typically stable enough for high-resolution SAS anyways, so it would require some sort of actively stabilized towfish anyways)
I’ve been thinking about, researching and doing some initial design work on this project for years but have been designing in earnest as of late and I finally am in a place where I can start working towards making this project a reality. This blog will jump into the middle of things, as there’s already so much work done that has gotten me to this stage. The main purpose of documenting it here is mainly as a record for myself, but if someone finds it interesting, that’s also great!
Some of the very high-level requirements I’ve set for this project:
Sonar resolution of better than 10cm in azimuth and range. (Goal: Better than 4cm)
The goal is driven by the physics of the design (transmitter dimensions and system bandwidth), but the actual requirement is loosened off due to some of the other challenges with forming SAS images.
A drawback to real aperture sonar such as side scan is that the azimuth resolution drops off with range, so even if this resolution is achievable at close ranges with a side scan, it is not possible across the entire range. With SAS, this resolution can be maintained across the entire image.
Sonar imaging range of at least 50 meters per side (Goal: 100 meters)
Driven by area mapping rate/speed and transmit power tradeoffs, as well as issues raised by the following requirement. Not too much else to say about this yet, other than it’s still in the trade space as I design the sonar. This may become configurable with different beam modes if I end up using a phased array transmitter.
Sonar imaging minimum depth/altitude of TBD meters
This is a challenging requirement to define, as shallow operation of sonar is very tough. Ideally, I’d set the minimums to something on the order of 10/5 meters, respectively, but I haven’t done enough analysis on the sonar yet to know if that’s achievable. This requirement will be refined as time goes on, so for now, the goal is “as shallow as possible”
AUV cruise velocity of up to 2 m/s through the water. Maximum mapping velocity of 1.5 m/s over ground
2 m/s should be enough to get through mild currents (while maintaining 1.5 m/s over ground). The flip side of this is that the propulsion design will need to have enough instantaneous thrust to keep the speed stable in spite of any environmental effects such as surge.
The 1.5 m/s stems from the preliminary sonar design and range requirements. The trade is between the sonar receive array length, range, and velocity. As previously stated, I may opt for configurable modes using a slower speed to achieve greater range.
AUV operational depth of 100 meters sea water. (Goal: 500 meters)
This is a nice round number, which some of the affordable design choices and material/component selections work well for. It’s also deeper than I’ll probably ever go in my scuba diving career, so even if I found something below that with the sonar, I wouldn’t be able to get to it!
The goal is very loose, would require upgrades (e.g. aluminum body sections as opposed to acrylic, better connectors, pressure compensation of the thruster, etc). The as-built design won’t meet the goal, but will be designed such that if I want to go down the upgrade route in the future, I wouldn’t have to start from scratch.
Likewise, the 100 meters won’t be the calculated buckling limit, but rather will have a very healthy safety margin attached to it — Basically it could go deeper, but gets riskier with any material/manufacturing imperfections causing failure.
2-way communications range of 1km
Both surface (RF) and subsurface (acoustic). Being a comms guy, this will be fun. I’ll probably use a COTS radio for surface communications as only a basic line-of-site datalink is needed, but the acoustic modem should be a fun challenge due to the strong multi-pathing. I’m tentatively aiming for at least 9.6 kbps rate communications, but will scale up as the component selections (transducer bandwidth, mainly) allows.
This is perhaps a poorly written requirement, but the aim is to make the AUV modular so that payloads can be swapped in and out. Potential future ideas are a camera payload, sensor payloads, translational thruster modules, etc. Once the basic requirements are met, I can expand to my heart’s content.
From those requirements, many other lower level requirements will come out, e.g. autonomous operations, mission duration, sonar frequencies, etc, etc. Many will flow out of the high level design process as I research and keep designing.
Ultimately, this is going to be a project that demands a lot of my limited free time, so the project may never get completed as time and resources limit how much I can do. The biggest requirement I’ve levied on myself is that I learn from this project — For my hobbies, the fun is in the journey, not just the destination.