Machine Learning Approach to Real-Time 3D Path Planning for Autonomous Navigation of Unmanned Aerial Vehicle

Abstract

The need for civilian use of Unmanned Aerial Vehicles (UAVs) has drastically increased in recent years. Their potential applications for civilian use include door-to-door package delivery, law enforcement, first aid, and emergency services in urban areas, which put the UAVs into obstacle collision risk. Therefore, UAVs are required to be equipped with sensors so as to acquire Artificial Intelligence (AI) to avoid potential risks during mission execution. The AI comes with intensive training of an on-board machine that is responsible to autonomously navigate the UAV. The training enables the UAV to develop humanoid perception of the environment it is to be navigating in. During the mission, this perception detects and localizes objects in the environment. It is based on this AI that this work proposes a real-time three-dimensional (3D) path planner that maneuvers the UAV towards destination through obstacle-free path. The proposed path planner has a heuristic sense of ☆A☆ algorithm, but requires no frontier nodes to be stored in a memory unlike ☆A☆. The planner relies on relative locations of detected objects (obstacles) and determines collision-free paths. This path planner is light-weight and hence a fast guidance method for real-time purposes. Its performance efficiency is proved through rigorous Software-In-The-Loop (SITL) simulations in constrained-environment and preliminary real flight tests.

Special Issue on Unmanned Aerial Vehicles

“…When moving in cluttered environments, collision-free reference trajectories are to be sought, to be followed by the vehicles. As path planing can be computationally demanding, a new light-weight planner is developed in [12] based on relative position of detected obstacles that can be used in real-time in a perception and control loop. Another approach is proposed in [13] that exploits obstacle geometry information to give priority to search in sub-spaces where a solution can be found quickly.…”

Coverage Path Planning with Semantic Segmentation for UAV in PV Plants

“…Tests of coverage paths it can be conducted with drones in a real or simulated environment. The most viable option for this stage of the work is simulation, as verified by other research [69][70][71][72]. Simulation has been recognized as an important research tool; initially, simulation was an academic research tool, but with the advancement of computers, simulation has reached new levels.…”

Unmanned aerial vehicles and machine learning for detecting objects in real time

“…It is therefore suggested that there are two practical but important restrictions to blame for this gap between application demands and technology capabilities [4]. It is difficult to build and store a variety of target object models, especially when the objects have a variety of appearances, and real-time object detection requires ISSN: 2302-9285  high computing power even to detect single objects, much less when many target objects are involved, in addition to object recognition algorithms that are tailored to specific object and context types [5].…”