Backseat Control of SandShark AUV using ROS on RaspberryPi

Abstract

This paper presents a backseat controller suitable for Autonomous Underwater Vehicles (AUVs) which promotes quick development turn-around of novel control methods, especially for missions which require onboard re-planning. This backseat controller communicates low-level control instructions to a vehicle-specific frontseat controller. Small, power efficient components comprise the backseat controller hardware, especially a Raspberry Pi computer. The backseat controller runs Robot Operating System (ROS), a middleware that facilitates implementation of novel control algorithms and use of OpenCV, an extensive collection of computer vision libraries. ROS enables the underwater community to leverage development efforts in the ground and aerial domains to accelerate progress. The backseat controller has been mated to General Dynamics Bluefin SandShark AUV. The platform has been validated in a controlled environment and is now being deployed in Lake Superior, MI, USA.