A compact exploration strategy for indoor flight vehicles

Abstract

This paper presents a compact exploration strategy designed to be implemented onboard indoor Miniature Air Vehicles (MAVs) operating in cluttered and confined environments. The exploration strategy uses 2D range information from a laser range scanner to generate velocity commands by blending a Sensor-based Random Tree frontier planner with a wall-following velocity field generator. The combined approach leverages the efficient exploration capabilities of frontier-based guidance and ensures that the vehicles follows a path that is free of obstacles and conducive to maintaining good scan geometry through a wall-following approach. The strategy has been successfully implemented and tested on a Quadrotor MAV and simulation results are presented. Flight test results shall be included in the final version.