Investigation of collaborative underwater vehicles for use of dynamic target tracking with communication considerations

Abstract

Underwater robotics has many practical applications for both government and industry, however, the environment poses distinct challenges for autonomous communication and perception of surroundings. A simulated environment and vehicles were created in the Robot Operating System (ROS). An experiment was designed to test three scenarios where AUVs working to track a dark ship are with various target speeds and mission behaviours. The three trials have increasing environmental and communication complexity, beginning in an ideal environment and concluding by operating in a virtual Bedford Basin, Canada. The autonomous vehicles rely on active sonar pings off the ship to receive range measurements and collaboratively trilaterate the location. Three motion primitives are defined and weighted to make eight mission types which characterize the movement of the AUVs. These are compared with a stationary profile to determine efficacy.