Autonomous Underwater Vehicle

Hello to everyone,

I’m trying to produce an autonomous underwater vehicle. I will do some work with image processing. I want to measure sound underwater and move towards the sound. I will use hydrophones for this, but I could not figure out how to make an algorithm. I thought of measuring the time difference between the sent signal and the received signal. I think I can find the distance by multiplying this time by the underwater speed.

Can you give an idea on this subject? What algorithm should I create?

Can I use hydrophones with Jetson nano?

How else can I measure underwater sound?

Thanks. :smiley:


I have a good amount of experience with underwater acoustic navigation for the Robosub competition. Could you describe what you’re trying to do a little better? Are you trying to do active or passive sonar - are you sending out pings and measuring reflections or just trying to measure a source in the water that you do not control.

Generally yes hydrophones are a good tool to use. You’ll either have to find one of these with an analog-to-digital converter attached, along with potentially a preamp and filter, or you will have to buy your own hydrophones and design a circuit that convert the analog voltage from the hydrophone into a digital signal, and potentially amplify and filter it. You will have to interface this with the Jetson Nano.

I’m not sure I follow what you want to do with image processing - do you have a sonar scan that you then want to do image processing on to determine where objects are? Or is it separate.


Thanks for your answer.

There is 1 pinger underwater. It will send a signal with a frequency of 45 kHz. I have to locate and destroy this pinger. I’m trying to make active sonar. I don’t know how to use the hydrophone with jetson nano. Which port will I connect to. Do I need to use analog data or digital data?

I have never worked with a hydrophone before, excuse me for this.

Do you mind if I ask what this is for and why you are trying to destroy a pinger, and how you actually plan on destroying the pinger from an AUV?

This might be a little presumptuous, but that seems like you are significantly overcomplicating the problem. If your target is actively sending out audio signals, then using passive sensors should be sufficient, and seems like it would be far simpler than trying to construct your own send-and-return sonar system. Using hydrophones (which look to be just microphones but designed for pressure waves in water instead of air), my background understanding is that you could set up an array of 4 and compare the timing of the signals to know exactly where the ping is coming from (time-difference-of-arrival or multilateration look like the keywords to search for). Alternatively, you could use a single directional sensor and move it around to find the “loudest” direction.

I don’t mean to be rude, but the questions you ask, as well as some of your other posts, suggest that you don’t have a lot of experience working with electromechanical systems. Jumping into a project like this with many new topics to learn may be more of a challenge than can be reasonably expected to be completed. It sounds like you have an objective and some good starting ideas, but not a clear design direction. If this is for a specific project with pre-determined tasks and constraints, knowing what those are and what you are expected to produce would be very helpful in trying to give you useful advice. If this is just an educational or side project, then I think taking more incremental steps that target a single specific aspect would serve you far more effectively. Your enthusiasm is heartening, and I certainly don’t want to squash it, but what you aiming for here might be too far a leap from where you are right now.


I agree @omerakblt. Do you already have the AUV built? This in and of itself is quite a large undertaking, and building an acoustics system on top of that is neither cheap nor easy, and will require lots of additional components beyond just a Jetson and hydrophone.

Gordon is correct that you will need an array of 4 hydrophones, no transmitters, and can compare time-difference-of-arrivals to determine where the signal is coming from.

I’m happy to provide more resources, I just want to know more about what this is for and that the resources are helpful with where you are in the project.

1 Like

Just throw this out there too. To help eliminate noise and interference, you might want to modulate your signals (i.e. put it on some sort of frequency like a 1kHz wave or something like that). Of course you’ll then need to measure your returning signal (capture the carrier wave). You can do this either completely through HW or SW.

Are you preparing for the upcoming water game?


I would start here as well.