Autonomous drone navigation methods review

Abstract

This paper provides a systematic review of self-navigation systems for drones (unmanned vehicles). The purpose of this review is to evaluate modern approaches to developing navigation strategies for autonomous vehicles. The data gathered and analyzed in this review can be used for developing new strategies for autonomous vehicle navigation without infringing intellectual property of original inventors, proposing well-informed solutions. This review is focused on analyzing methodologies described in 140 technological patents filed in the last several years. The patents for the review were gathered from specialized databases and Google Patents. The results of this review were synthesized by comparing the hardware (sensor type, controller type, etc.) and software (system, algorithms) described in the implementation section of a patent. Because patent information can only give a rough idea of existing industrial trends, and actual software and hardware used by a manufacturer is typically a trade secret or protected procedural knowledge, this review’s synthesized results are only partially complete. However, the results provide some insights into the ways drone navigation systems are currently developed and highlights the most widely used technology (e.g. the prevalence of machine learning algorithms for pattern recognition) and methods (e.g. fuzzy logic).