The range camera in Microsoft's Kinect, intended for the Xbox 360 gaming console, offers a powerful alternative to the many standard sensors used in robotics for gathering spatial information about a robot’s surroundings. The recently-released Kinect is the first commercially available product to provide depth data of its resolution and accuracy with a price tag within reach of many robotics projects. The work described in this paper explores the feasibility of using this sensor by developing a robot that relies solely on the Kinect for sensory data. This robot successfully performs standard navigational procedures, demonstrating the possibility of integrating spatial information from the Kinect into a real-time robotics application. This paper documents the techniques used to integrate the Kinect into the system, highlighting the key benefits and limitations of the sensor. / text

Unmanned ground vehicles (UGVs) have been widely used in many areas such as agriculture, mining, construction and military applications. This results from the fact that UGVs can not only be easily built and controlled, but also be featured with high mobility and handling hazardous situations in complex environments. Among the competences of UGVs, autonomous navigation is one of the most challenging problems. This is because that the success in achieving autonomous navigation depends on four factors: Perception, localization, cognition, and proper motion controller. In this thesis, we introduce the realization of autonomous navigation for a two-wheeled differential ground robot under the robot operating system (ROS) environment from both the simulation and experimental perspectives. In Chapter 2, the simulation work is discussed. Firstly, the robot model is described in the unified robot description format (URDF)-based form and the working environment for the robot is simulated. Then we use the \textit{gmapping} package which is one of the packages integrating simultaneous localization and mapping (SLAM) algorithm to build the map of the working environment. In addition, ROS packages including \textit{tf}, \textit{move\_base}, \textit{amcl}, etc., are used to realize the autonomous navigation. Finally, simulation results show the feasibility and effectiveness of the autonomous navigation system for the two-wheeled UGV with the ability to avoid collisions with obstacles. In Chapter 3, we introduce the experimental studies of implementing autonomous navigation for a two-wheeled UGV. The necessary hardware peripherals on the UGV to achieve autonomous navigation are given. The process of implementation in the experiment is similar to that in simulation, however, calibration of several devices is necessary to adapt the scenario in a practical environment. Additionally, a proportional-integral-derivative (PID) controller for the robot base is used to handle the external noise during the experiment. The experimental results demonstrate the success in the implementation of autonomous navigation for the UGV in practice. / Graduate

The use of potential field based navigation schemes in robotics has been limited by inherent local minima issues. Local minima traps, small passages, unstable motion, and targets positioned near objects all pose major concerns when using potential fields for local vehicle control. This work proposes a new algorithm, "Re-Active Vector Equilibrium" (RAVE) that mitigates many of these issues. The vehicle representation model is expanded to use multiple points subject to potential calculation and the addition of two forces, a velocity dependent "risk force" (F_rsk) and a velocity and direction dependent "tangential force" (F_tan). The vehicle representation model is also expanded from a single reactive point to a series of points that define the vehicle body, providing better and simpler vehicle control. This has the effect of simplifying the required calculations at the cost of increasing the calculation count. The risk force, F_rsk, allows for dynamic adaptation to the immediate environment by acting in opposition to the net obstacle force, and is inversely proportional to the vehicle speed. The tangential force, F_tan, encourages better wall-following behaviour and provides a biasing mechanism to resolve obstacle aligned with target local minima issues.

<div>We describe a new task called The Amazing Race: Robot Edition. In this task, the robot is placed in a real, unknown environment, without a map, and asked to find a designated location. It will need to explore its surroundings, find and approach people, engage them in a dialogue to obtain directions to the goal, and follow those directions to the hallway with the goal. We describe and implement a variety of robotic behaviors that performs each of these functions. We test these in the real world in test environments that were distinct from the training environments where we developed our methods and trained our models. Additionally, these test environments were completely unmodified and reflect the state of the real world.</div><div>First, we describe how our robotic system solves this problem where the environment is constrained to a single floor or a single building. We demonstrate that we are able to find a goal location in never-before-seen environments. Next, we describe a machine-learned approach to the dialogue and components of our system to make it more robust to the diversity and noisiness of navigational instructions someone may provide.</div>

There are two commonly accepted paradigms for organizing intelligence in robotic vehicles, namely reactive and deliberative. A third, a hybrid paradigm called integrated planning and execution, is considered a combination of the original two. Although these paradigms are well known to researchers, there are few published examples directly comparing their application and performance on similar vehicles operating in identical environments. Virginia Tech's participation with two nearly identical vehicles in the DARPA Grand Challenge afforded a practical opportunity for such a case study. Both base vehicles were developed by modifying Club Car Pioneer XRT 1500 on-demand four wheel drive base platforms. Cliff was designed to use the reactive paradigm, while Rocky was designed to use the deliberative paradigm. Both vehicles were initially outfitted with sensor suites and computational capabilities commensurate with the paradigm being employed. The author of this thesis coordinated the activities of the two teams of undergraduate and graduate students who implemented the respective designs and software. Both vehicles proved capable of off-road navigation, including road following and obstacle avoidance in complex desert terrain. In the end, however, the reactive paradigm proved to be smoother and more reliable than the deliberative paradigm under the conditions of our testing. While both vehicles were extensively tested and compared using the competing paradigms, the team modified Rocky to use the more effective reactive paradigm for the Grand Challenge events. The deliberative case shows much promise for complex navigation, but added unnecessary complexity to desert road navigation. This case study, while necessarily limited in scope, may help to shed additional light on the tradeoffs and performance of competing approaches to machine intelligence. / Master of Science

No description available.

An issue of concern for maritime safety when operating a small to medium-sized sailboat is that the presence of hazards in the navigational route in the form of floating logs can lead to a severe collision if undetected. As a precautionary measure to prevent such a collision with a log, a 2D vision-based detection system is proposed. We take a combined approach involving predictive mapping by linear regression and saliency detection. This approach is found to overcome specific issues related to the illumination changes and unstructured environment in the dataset. The proposed method has been evaluated using precision and recall measures. This proof of concept demonstrates the potential of the method for deployment on a real-time onboard detection system. The algorithm is robust and of reasonable computational complexity. / Graduate

FundaÃÃo Cearense de Apoio ao Desenvolvimento Cientifico e TecnolÃgico / Sistemas de navegaÃÃo autÃnomos devem ser capazes de definir uma seqÃÃncia de aÃÃes a serem tomadas por robÃs mÃveis, dotados de um conjunto limitado de sensores, quando expostos a um ambiente externo suposto desconhecido e tendo que atender simultaneamente a um elenco de objetivos previamente especificados. O interesse cientÃfico no estudo de sistemas de navegaÃÃo em ambientes desconhecidos e sujeitos ao atendimento de mÃltiplos objetivos à motivado basicamente pelo seu evidente potencial em aplicaÃÃes industriais e pelo fato de demandarem a implementaÃÃo de estratÃgias de soluÃÃes complexas. Este trabalho apresenta a modelagem de um sistema de navegaÃÃo para robÃs moveis por meio de Redes de Petri Coloridas. O modelo apresentado consegue simular vÃrias situaÃÃes, tais como a representaÃÃo do mundo em volta do robÃ, interaÃÃo com o ambiente, planejamento de trajetÃria, localizaÃÃo do robà e anÃlise das baterias, bem como servir de base para implementaÃÃo em um robà mÃvel real e otimizaÃÃo do sistema. / Systems of autonomous navigation must be able to define a sequence of actions to be taken bymobile robots endowed with a set of limited sensors, while exposed to an unknown environment and having to serve simultaneously a set of objectives previously specified. The scientific interest in the study of systems of navigation in unknown environment which are subject of serving to several objectives is motivated basically by its evident potential in industrial applications and by the fact of demanding the implementation of complex solutions strategies. This research presents the modeling of a navigation system for mobile robots through coloured Petri nets. The model presented here can simulate several situations, such as: the representation of the world around the robot, interaction with the environment, trajectory planning, robot location, battery analysis, as well as how to serve as a basis for implementation in a real mobile robot and optimization of the system.

<p>This thesis describes an ongoing project of which the purpose is designing and developing techniques and algorithms for autonomous off-road vehicles. The focus is on some of the components necessary to accomplish autonomous navigation, which involves sensing and moving safely along a user-defined path in a dynamic forest environment. The work is part of a long-term vision in the forest industry of developing an unmanned shuttle that transports timber from the felling area to the main roads for further transportation. A new path-tracking algorithm is introduced and demonstrated as superior to standard algorithms, such as Follow the Carrot and Pure Pursuit. This is accomplished by using recorded data from a path-learning phase. By using the recorded steering angle, the curvature of the path is automatically included in the final steering command. Localization is accomplished by a neural network that fuses data from a Real-Time Kinematic Differential GPS/GLONASS, a gyro, and wheel odometry. Test results are presented for path tracking and localization accuracy from runs conducted on a full-sized forest machine. A large part of the work has been design and implementation of a general software platform for research in autonomous vehicles. The developed algorithms and software have been implemented and tested on a full-size forest machine supplied by our industrial partner Komatsu Forest AB. Results from successful field tests with autonomous path tracking, including obstacle avoidance, are presented.</p>

Mobile robotics research is currently interested on exporting autonomous navigation results achieved in indoor environments, to more challenging environments, such as, for instance, urban pedestrian areas. Developing mobile robots with autonomous navigation capabilities in such urban environments supposes a basic requirement for a upperlevel service set that could be provided to an users community. However, exporting indoor techniques to outdoor urban pedestrian scenarios is not evident due to the larger size of the environment, the dynamism of the scene due to pedestrians and other moving obstacles, the sunlight conditions, and the high presence of three dimensional elements such as ramps, steps, curbs or holes. Moreover, GPS-based mobile robot localization has demonstrated insufficient performance for robust long-term navigation in urban environments. One of the key modules within autonomous navigation is localization. If localization supposes an a priori map, even if it is not a complete model of the environment, localization is called map-based. This assumption is realistic since current trends of city councils are on building precise maps of their cities, specially of the most interesting places such as city downtowns. Having robots localized within a map allows for a high-level planning and monitoring, so that robots can achieve goal points expressed on the map, by following in a deliberative way a previously planned route. This thesis deals with the mobile robot map-based localization issue in urban pedestrian areas. The thesis approach uses the particle filter algorithm, a well-known and widely used probabilistic and recursive method for data fusion and state estimation. The main contributions of the thesis are divided on four aspects: (1) long-term experiments of mobile robot 2D and 3D position tracking in real urban pedestrian scenarios within a full autonomous navigation framework, (2) developing a fast and accurate technique to compute on-line range observation models in 3D environments, a basic step required by the real-time performance of the developed particle filter, (3) formulation of a particle filter that integrates asynchronous data streams and (4) a theoretical proposal to solve the global localization problem in an active and cooperative way, defining cooperation as either information sharing among the robots or planning joint actions to solve a common goal. / Actualment, la recerca en robòtica mòbil té un interés creixent en exportar els resultats de navegació autònoma aconseguits en entorns interiors cap a d'altres tipus d'entorns més exigents, com, per exemple, les àrees urbanes peatonals. Desenvolupar capacitats de navegació autònoma en aquests entorns urbans és un requisit bàsic per poder proporcionar un conjunt de serveis de més alt nivell a una comunitat d'usuaris. Malgrat tot, exportar les tècniques d'interiors cap a entorns exteriors peatonals no és evident, a causa de la major dimensió de l'entorn, del dinamisme de l'escena provocada pels peatons i per altres obstacles en moviment, de la resposta de certs sensors a la il.luminació natural, i de la constant presència d'elements tridimensionals tals com rampes, escales, voreres o forats. D'altra banda, la localització de robots mòbils basada en GPS ha demostrat uns resultats insuficients de cara a una navegació robusta i de llarga durada en entorns urbans. Una de les peces clau en la navegació autònoma és la localització. En el cas que la localització consideri un mapa conegut a priori, encara que no sigui un model complet de l'entorn, parlem d'una localització basada en un mapa. Aquesta assumpció és realista ja que la tendència actual de les administracions locals és de construir mapes precisos de les ciutats, especialment dels llocs d'interés tals com les zones més cèntriques. El fet de tenir els robots localitzats en un mapa permet una planificació i una monitorització d'alt nivell, i així els robots poden arribar a destinacions indicades sobre el mapa, tot seguint de forma deliberativa una ruta prèviament planificada. Aquesta tesi tracta el tema de la localització de robots mòbils, basada en un mapa i per entorns urbans peatonals. La proposta de la tesi utilitza el filtre de partícules, un mètode probabilístic i recursiu, ben conegut i àmpliament utilitzat per la fusió de dades i l'estimació d'estats. Les principals contribucions de la tesi queden dividides en quatre aspectes: (1) experimentació de llarga durada del seguiment de la posició, tant en 2D com en 3D, d'un robot mòbil en entorns urbans reals, en el context de la navegació autònoma, (2) desenvolupament d'una tècnica ràpida i precisa per calcular en temps d'execució els models d'observació de distàncies en entorns 3D, un requisit bàsic pel rendiment del filtre de partícules a temps real, (3) formulació d'un filtre de partícules que integra conjunts de dades asíncrones i (4) proposta teòrica per solucionar la localització global d'una manera activa i cooperativa, entenent la cooperació com el fet de compartir informació, o bé com el de planificar accions conjuntes per solucionar un objectiu comú.