What I want to do or know
I want to know if the Thermal Fishing Bob concept could work as an Autonomous "SeaGlider" type robot.
A "SeaGlider" is a type of autonomous underwater vehicle that mimics a "porpoising" motion to propel itself through the water. The method uses far less energy than a conventional propeller-driven method and is being researched for long endurance vehicles.
An impressive, low-cost, DIY version called the SeaGlide has been developed as a school curriculum by the same organization that sponsored the (far less impressive IMHO) "Sea Perch" competition. The web site at http://www.seaglide.net/ includes a full lesson plan, BOM, 3D printable part files and Arduino firmware. All of which can be downloaded freely from this Dropbox-https://www.dropbox.com/sh/lattz1cupiqa6ai/AAD40WVk4Fu62y14ojnYm_D1a?dl=0 The fully functioning unit could be built for less than $100 depending on the availability of materials.
The SeaGlide already includes a "Sensor Pod" with an Arduino, temp sensor, pressure and even an RGB LED to indicate the vehicles water depth. So the vehicle could function in a manner similar to a Thermal Fishing Bob in its current configuration. The only issue is whether or not the diving motion would make a long exposure image that accurately reflects water temperature too difficult to capture. If it were possible to account for water temperature changes at different depths, then I imagine the diving motion would be an advantage, but I'm not sure how this would be accomplished.
Of course, the advantages of such a low-cost, autonomous vehicle could make it appealing for a wide variety of uses beyond thermal imaging. The arduino "sensor pod" allows for all kinds of device configurations...
I've wanted to build one myself for some time, but living as I do on the ocean, the vehicle is really more appropriate for rivers, ponds or other inland waterways where you could keep it from swimming away!