Autonomous Navigation of a Robotic Swarm in Space Exploration Missions

Abstract

In recent years the paradigm of navigation has shifted from pinpointing the location of a single agent to continuously estimating the full kinematic state of a team of autonomous agents. In this paper, we propose a kinematic-aware information seeking algorithm for a robotic swarm. The algorithm tightly couples state estimation and autonomous control given ranging and kinematic models. With the help of the Fisher information theory, agents generate information seeking command sequence on their actuators, which leads to smooth trajectories. As an outcome, the swarm continuously optimizes its formation so that the agents’ position and orientation uncertainty is actively minimized. The proposed algorithm is verified by physics simulations and demonstrated in a space-analog mission of autonomous swarm navigation on volcano Mount Etna.