A short update on my whale song project: it’s shifting technology.
I had a nice exchange with the CTO of Sofar. There are patents on the connectors. They also do own the data that moves across their network. They have a reason for that, and I respect that. It just does not work for me.
I’ve also learned a ton. I don’t think cellular or satellite data is needed for my project. I will be in coastal/bay waters well within link range for several other technologies. I also don’t think that transferring full whale song audio over those links is yet commercially viable.
I’ve discovered that many scientists have literally thousands of hours of whale song recording. We need to move forward with ML on that. Scientist are making breakthroughs on decoding whale song, including mapping an alphabet of sperm whale clicks. The problem isn’t “we need to record more whale songs” it’s “we need to decode it, and then we need ways to replay it.” This means we need to focus on an easy way to have transponders to not just receive but also to transmit. And the sensor package in the water will need more power than what the SoFar can do. So the Spotter looks less and less like a good fit for my project.
So, I’m returning the Sofar Spotter, as cool as it is. Their CTO understands my position and has offered me some node boards/etc. to test with if I need them. This lets someone else have this dev kit and do something cool, and I’ll stay connected to the company and team.
My whale song efforts are very much a side project for me, but very much on my mind. I’m still in awe of these majestic creatures of the deep and want to apply my technology skills to help understand them!
Stay tuned. Expect some radio technology discussion, some more Linux, some ML and perhaps even CUDA. And some acoustic transponder stuff too, soon.
