When I started hatching this project in September 2015, I swore I wouldn't fool myself about how much time, money, work, and disappointment I would have to endure to complete this project. It's ambitious, but I believe that, with my professional robotics background and with the abilities of all the great people that will step forward to contribute, we can successfully deploy an AUV to explore the ocean floor around NYC.
To this end, I've brooded over the shortest and most fulfilling path to successful deployment that doesn't break my wallet. From the brooding emerged some key decisions:
- I purchased a BlueROV from Blue Robotics. I had planned to buy thrusters from the company regardless, but now with the addition of a basic hull, we need not worry about mechanical designs and trial-and-error fabrication processes. I'd love to have a custom, more fluidly-dynamic hull, but it's not viable for someone with my limited mechanical knowledge to design that.
- The AUV needs a positioning system. A Doppler Velocity Log (DVL) is too expensive ($25k+). While IMUs might afford it an acceptable dead-reckoned position, the models capable of that are military grade and have military price tags. A hydrophone array which triangulates its position to a subsurface acoustic pinger will costs ~$5000 (that's good), and will provide an accurate position relative to the subsurface pinger. Such behavior is fine for our use cases.
- I purchased three single board computers from ODroid, which will operate three different on-board systems-- High-level behavior + image processing, navigation, and safety. ODroid makes nice SBCs for which they've created Ubuntu server ISOs.
All other major decision remain in flux, and the not-knowing is always exciting.