Components

Depth Control - Vertical Thruster

    A thruster mounted near the center of mass of the submersible with a vertical axis of rotation will provide vertical thrust to maintain and change depth.  A closed-loop controller will vary the vertical thrust to maintain a set distance off the ocean floor.  Constant downward thrust will be required to maintain depth, counteracting the passive buoyancy of the submersible.

Propulsion - OpenROV Thrusters

    OpenROV's brushless DC motors and accompanying propellors, the same as those found on the OpenROV 2.8 tethered rover, have been selected for the main propulsion system of this submersible.  Two thrusters will be placed toward the aft of the submersible, flanking the electronics enclosure, to provide yaw authority by applying a differential of thrust between the two thrusters.

Custom Enclosure - Machined Polycarbonate

Machining Electronics Housing: Center Duct (Left), Bonding(Right)

    Custom enclosure for the electronics and center thruster are built out of machined polycarbonate and glued together using solvent weld to form boxes.  These custom enclosures will be the minimum size necessary to comfortably fit the required electronics within the submersible.  Laser-cut gaskets will be utilized to seal access hatches.  To ensure the safety of the electronics, the enclosures will be pressure-tested up to 1 atm before any electronics will be fitted.

Laser Cut Neoprene Gaskets

Imaging - Raspberry Pi Camera Module v2

Range Finding - Laser Triangulation

Laser Triangulation Range Finder 

    The distance to the ocean floor will be determined by identifying the location of a laser dot in images captured by the down-facing Raspberry Pi Camera Module v2 used for mapping.  The laser dot will be produced by a 5 mW red laser diode pointing down and parallel to the camera's orientation.  Thus as the submersible moves further from the ocean floor, the laser dot will appear closer to the center of the captured image as seen in the figure.  Detection of the laser dot will be performed in OpenCV and pixel location is mapped to corresponding range.  When not being used to detect range, the laser will be turned off so that it does not appear in the map generated by the submersible.

Laser Triangulation Mathematical Model

Localization - Feature Tracking

    The submersible will maintain an estimate of its position by analyzing images captured by its downward-facing camera.  Features will be identified and matched between the two most recent images in real-time.  If the distance to the ocean floor is known, then the average displacement of these features can be used to determine how far the submersible moved.  The sum of all these small displacements, combined with knowledge of the submersible's heading, will then yield a reasonably accurate estimate of the submersible's position.