ROS-based simulation environment for obstacle avoidance in autonomous navigation

Abstract

In this paper, we present a Robot Operating System (ROS) based simulation environment for obstacle avoidance in maritime operations, suitable for different planning algorithms. The chosen algorithm in the current implementation is the Timed Elastic Band (TEB) available in the TEB_local_planner package of ROS. It allows to plan paths that are consistent with the kinematic constraints of an Autonomous Surface Vehicle (ASV) and to optimize the planning according to several parameters such as the navigation time and the Cross Track Error. The proposed framework has been validated and its performance assessed via Monte Carlo simulations.