This is the second ROV (remotely operated vehicle) my high school built for the MATE competition. This was our second year participating and our robot definitely had a few extra tricks up its sleeve compared to last year’s ROV. First of all, the frame was built out of aluminum compared to the heavier PVC pipe used the previous year. This made the robot more maneuverable underwater. Another upgrade was the bilge pump configuration, which consisted of six independently controlled bilge pumps: two were used to move forward, back, and rotate, and the other four were used to move up, down, and control the pitch of the ROV. To control the pitch of the ROV meant that the ROV had to be equally weighted underwater, unlike last years ROV which had ballasts on the top and weights at the bottom. The ballasts provided a force up, and the weights a force down, keeping the ROV in the up-right orientation. So to control the ROV’s pitch this year, the ROV was weighted with a center of gravity (underwater) in the middle.

Another major makeover from last year was the electronics enclosure. Last year an electronics box was concealed in wax to prevent the water from leaking in. Not only did the water leak through the wax, the electronics were now stuck in a wax block that had to be melted to access (we spent three hours after the competition with a blow dryer doing just that). So this year we designed an accessible chamber for the electronics. The electronics were accessible through, the camera holder on the front by twisting it off. Teflon tape was used to seal the threads. The electronics board was designed to slide out of the PVC, giving complete access to all the circuits. The wires going to and from the electronics (for power, bilge pumps, sensors, solenoid on claw, and cat5 communications line) were all routed out two holes in the back which were caulked to keep water out.

This year’s ROV only used one camera compared to two because this year’s tasks only required one camera. The camera was mounted in the same enclosure as the electronics which prevented having three wires (two for power and one for data) going from the electronics box to an external camera enclosure. This was the case last year, and one of the camera enclosures leaked.

The claw was also rebuilt. Last year a Lego claw was built which wasn’t capable of performing many of the tasks (I love Legos and their versatility, but there’s an extent to when they shouldn’t be used). This year we stepped it up and built a claw out of aluminum. We milled custom parts and assembled the claw as designed. Powering the claw was a solenoid which proved to be more reliable than the geared motor from last year. A weak spring held the claw open until a current passed through the solenoid, tightly pushing the claw together.

The electronics were designed onto a PCB and a new microcontroller was used. The Propeller Board was replaced with the Arduino Duemilanove and the motor driver circuitry was designed onto a PCB in Express PCB. The PCB layout can be downloaded here.

The Arduino was programmed in C using the provided IDE. The Arduino stored incoming serial data on its 128 MB buffer and the Arduino’s built in ‘SerialRead()’ function was used to retrieve the data. The code then processed the data and adjusted the motors accordingly. The latest code for the Arduino can be found here.

On the computer side, a C++ program read the state of an Xbox 360 controller and sent the serial data to the Arduino. To read the state of the Xbox controller the XInput API was used. The C++ code can be downloaded here.