Introduction
One of the several projects I undertook at MIT for the academic year 2005-2006 was an ROV for the class 2.011: Introduction to Ocean Science and Engineering. The design and construction of the ROV, shared between the students taking the class, formed the practical coursework requirement; for more details on the class (including more pictures of the ROV and a video of its field trials), see the OpenCourseWare website for the Spring 2006 offering.
The ROV we built was dubbed the "MegaSeaPerch" - it was based on the "SuperSeaPerch" ROV developed by MIT's Sea Grant Laboratory, which in turn was descended from the "SeaPerch" ROV used for the yearly "Discover Ocean Engineering" classes.
I and a number of others worked specifically on the control system for the vehicle, in my case thanks to my background in electrical engineering. We developed the controls in two phases: Phase I was a simple switch-box controller of the ROV's 4 thrusters, while Phase II involved computer control with a Tattletale 8 microcontroller onboard the vehicle communicating over a serial connection to a laptop running a Java GUI/joystick interface. The Tattletale controlled the motors through off-the-shelf radio-control-modelling speed controllers running at 24V, and also received data from the vehicle's onboard sensors (for pressure, temperature and water speed).
The finale of the class was the testing of the ROV in the sea at Woods Hole, Massachusetts (site of the Woods Hole Oceanographic Institute. We all had great fun, although (for some reason best none to itself) the onboard microcontroller did not sample any of the sensor readings; it was just quite amazing to see something that we had built, mostly at breakneck speed, actually working! (Certainly I was most surprised that the electronics and software I had hacked together didn't fail in some obscure way - the wonders of systems engineering!) See also the video of its field trials.
More technical details
Project details and links
