In this paper, we present a motion planning framework for a fully deployed autonomous unmanned aerial vehicle which integrates two sample-based motion planning techniques, Probabilistic Roadmaps and Rapidly Exploring Random Trees. Additionally, we incorporate dynamic reconfigurability into the framework by integrating the motion planners with the control kernel of the UAV in a novel manner with little modification to the original algorithms. The framework has been verified through simulation and in actual flight. Empirical results show that these techniques used with such a framework offer a surprisingly efficient method for dynamically reconfiguring a motion plan based on unforeseen contingencies which may arise during the execution of a plan. The framework is generic and can be used for additional platforms.

The use of Unmanned Aerial Vehicles (UAVs) which can operate autonomously in dynamic and complex operational environments is becoming increasingly more common. While the application domains in which they are currently used are still predominantly military in nature, in the future we can expect wide spread usage in thecivil and commercial sectors. In order to insert such vehicles into commercial airspace, it is inherently important that these vehicles can generate collision-free motion plans and also be able to modify such plans during theirexecution in order to deal with contingencies which arise during the course of operation. In this paper, wepresent a fully deployed autonomous unmanned aerial vehicle, based on a Yamaha RMAX helicopter, whichis capable of navigation in urban environments. We describe a motion planning framework which integrates two sample-based motion planning techniques, Probabilistic Roadmaps and Rapidly Exploring Random Treestogether with a path following controller that is used during path execution. Integrating deliberative services, suchas planners, seamlessly with control components in autonomous architectures is currently one of the major open problems in robotics research. We show how the integration between the motion planning framework and thecontrol kernel is done in our system.

Additionally, we incorporate a dynamic path reconfigurability scheme. It offers a surprisingly efficient method for dynamic replanning of a motion plan based on unforeseen contingencies which may arise during the execution of a plan. Those contingencies can be inserted via ground operator/UAV interaction to dynamically change UAV flight paths on the fly. The system has been verified through simulation and in actual flight. We present empirical results of the performance of the framework and the path following controller.

Unmanned aircraft systems use a variety of techniques to plan collision-free flight paths given a map of obstacles and no- fly zones. However, maps are not perfect and obstacles may change over time or be detected during flight, which may in- validate paths that the aircraft is already following. Thus, dynamic in-flight replanning is required.Numerous strategies can be used for replanning, where the time requirements and the plan quality associated with each strategy depend on the environment around the original flight path. In this paper, we investigate the use of machine learn- ing techniques, in particular support vector machines, to choose the best possible replanning strategy depending on the amount of time available. The system has been implemented, integrated and tested in hardware-in-the-loop simulation with a Yamaha RMAX helicopter platform.

This paper presents a software framework which combines reactive collision avoidance control approach with path planning techniques for the purpose of safe navigation of multiple Unmanned Aerial Vehicles (UAVs) operating in unknown environments. The system proposed leverages advantages of using a fast local sense-and-react type control which guarantees real-time execution with computationally demanding path planning algorithms which generate globally optimal plans. A number of probabilistic path planning algorithms based on Probabilistic Roadmaps and Rapidly-Exploring Random Trees have been integrated. Additionally, the system uses a reactive controller based on Optimal Reciprocal Collision Avoidance (ORCA) for path execution and fast sense-and-avoid behavior. During the mission execution a 3D map representation of the environment is build incrementally and used for path planning. A prototype implementation on a small scale quad-rotor platform has been developed. The UAV used in the experiments was equipped with a structured-light depth sensor to obtain information about the environment in form of occupancy grid map. The system has been tested in a number of simulated missions as well as in real flights and the results of the evaluations are presented.

This paper presents the design and development of the LinkBoard, an advanced flight control system for micro Unmanned Aerial Vehicles (UAVs). Both hardware and software architectures are presented. The LinkBoard includes four processing units and a full inertial measurement unit. In the basic configuration, the software architecture includes a fully configurable set of control modes and sensor fusion algorithms for autonomous UAV operation. The system proposed allows for easy integration with new platforms, additional external sensors and a flexibility to trade off computational power, weight and power consumption. Due to the available onboard computational power, it has been used for computationally demanding applications such as the implementation of an autonomous indoor vision-based navigation system with all computations performed onboard. The autopilot has been manufactured and deployed on multiple UAVs. Examples of UAV systems built with the LinkBoard and their applications are presented, as well as an in-flight experimental performance evaluation of a newly developed attitude estimation filter.

Due to the maturity of technological development, widespread use of Unmanned Aerial Vehicles (UAVs) is becoming prevalent in the civil and commercial sectors. One promising area of application is in emergency rescue support. As recently seen in a number of natural catastrophes such as the hurricanes in Texas, Florida and Puerto Rico, major communication and electrical infrastructure is knocked out, leading to an inability to communicate between the victims and rescuers on the ground as well as between rescuers themselves. This paper studies the feasibility of using heterogeneous teams of UAVs to rapidly deliver and establish ad hoc communication networks in operational environments through autonomous in-air delivery of CommKits that serve as nodes in local ad hoc networks. Hardware and software infrastructures for autonomous CommKit delivery in addition to CommKit specification and construction is considered. The results of initial evaluation of two design alternatives for CommKits are presented based on more than 25 real flight tests in different weather conditions using a commercial small-scale UAV platform.

The focus of this paper is on base functionalities required for UAV-based rapid deployment of an ad hoc communication infrastructure in the initial phases of rescue operations. The main idea is to use heterogeneous teams of UAVs to deploy communication kits that include routers, and are used in the generation of ad hoc Wireless Mesh Networks (WMN). Several fundamental problems are considered and algorithms are proposed to solve these problems. The Router Node Placement problem (RNP) and a generalization of it that takes into account additional constraints arising in actual field usage is considered first. The RNP problem tries to determine how to optimally place routers in a WMN. A new algorithm, the RRT-WMN algorithm, is proposed to solve this problem. It is based in part on a novel use of the Rapidly Exploring Random Trees (RRT) algorithm used in motion planning. A comparative empirical evaluation between the RRT-WMN algorithm and existing techniques such as the Covariance Matrix Adaptation Evolution Strategy (CMA-ES) and Particle Swarm Optimization (PSO), shows that the RRT-WMN algorithm has far better performance both in amount of time taken and regional coverage as the generalized RNP problem scales to realistic scenarios. The Gateway Node Placement Problem (GNP) tries to determine how to locate a minimal number of gateway nodes in a WMN backbone network while satisfying a number of Quality of Service (QoS) constraints.Two alternatives are proposed for solving the combined RNP-GNP problem. The first approach combines the RRT-WMN algorithm with a preexisting graph clustering algorithm. The second approach, WMNbyAreaDecomposition, proposes a novel divide-and-conquer algorithm that recursively partitions a target deployment area into a set of disjoint regions, thus creating a number of simpler RNP problems that are then solved concurrently. Both algorithms are evaluated on real-world GIS models of different size and complexity. WMNbyAreaDecomposition is shown to outperform existing algorithms using 73% to 92% fewer router nodes while at the same time satisfying all QoS requirements.