Search

This work presents real time implementation algorithms of Simultaneous Localization and Map Building (SLAM) with emphasis to outdoor land vehicle applications in large environments. It presents the problematic of outdoors navigation in areas with combination of feature and featureless regions. The aspect of feature detection and validation is investigated to reliably detect the predominant features in the environment. Aided SLAM algorithms are presented that incorporate absolute information in a consistent manner. The SLAM implementation uses the compressed filter algorithm to maintain the map with a cost proportional to number of landmarks in the local area. The information gathered in the local area requires a full SLAM update when the vehicle leaves the local area. Algorithms to reduce the full update computational cost are also presented. Finally, experimental results obtained with a standard vehicle running in unstructured outdoor environment are presented. [ABSTRACT FROM AUTHOR]

Knowing its position in an environment is an essential capability for any useful mobile robot. Monte Carlo Localization (MCL) has become a popular framework for solving the self-localization problem in mobile robots. The known methods exploit sensor data obtained from laser range finders or sonar rings to estimate robot positions and are quite reliable and robust against noise. An open question is whether comparable localization performance can be achieved using only camera images, especially if the camera images are used both for localization and object recognition. In such a situation, it is both harder to obtain suitable models for predicting sensor readings and to correlate actual with predicted sensor data. Both problems can be easily solved if localization is haiti on features obtained by preprocessing images. In our image-based MCL approach, we combine visual distance features and visual landmark features, which haw different characteristics. Distance features can always be computed, but have value-dependent and dramatically increasing margins for noise. Landmark features give good directional information, but are detected only sporadically. In our paper, we discuss the problems arising from these characteristics and show experimentally that MCL nevertheless works very well under these conditions. [ABSTRACT FROM AUTHOR]

Topology-based maps are a new representation of the workspace of a mobile robot, which capture the structure of the flee space in the environment in terms of the basic topological notions of connectivity and adjacency. A topology-based map can represent the environment terms of open spaces (rooms and corridors) connected by narrow passages (doors and junctions). In this paper, we show how to enrich a topology-based map with geometric information useful for the generation and execution of navigation plans. Both the topology-based map and its geometric information are automatically extracted from sensor data. We illustrate the use of topology-based maps for planned behavior-based navigation on a real robot. [ABSTRACT FROM AUTHOR]

We have to prepare the evaluation (fitness) function to evaluate the performance of the robot when we apply the machine learning techniques to the robot application. In many cases, the fitness function is composed of several aspects. Simple implementation to cope with the multiple fitness functions is a weighted summation. This paper presents an adaptive fitness function for the evolutionary computation to obtain the purposive behaviors through changing the weights for the fitness function. As an example task, a basic behavior in a simplified soccer game (shooting a ball into the opponent goal) is selected to show the validity of the adaptive fitness function. Simulation results and real experiments are shown, and a discussion is given. [ABSTRACT FROM AUTHOR]

Presents a self-localization method based on Hough transformation for mobile robots in polygonal environments. Techniques for robot self-localization; Problem in mobile robotics; Aspects of robotic techniques based on design choices.

Develops a path tracking scheme for a car-like mobile robot based on neural predictive control. Employment of a multi-layer back-propagation neural network; Formation of wavelet neural network; Provision of constructive ways to build up neural networks.

Describes the design of a practical system using natural language to teach a mobile robot in navigating a miniature town. Provision of primitive procedures derived from a corpus of instructions; Components of the instruction-based learning architecture; Role played by the dialogue management.

Focuses on the way a robot can autonomously learn the execution of navigation plans. Use of a Markov Decision Process; Application of the neural networks and regression trees; Usage of autonomous mobile robots to service tasks.

Examines the visual approach skill for a mobile robot using learning and fusion of simple skills. Advantage of the reinforcement learning algorithm; Implementation of perceptive and sensorimotor skills in a robot; Adaptation of changes in the environment.

Describes a framework combining game theory and geometry used in addressing target-tracking problem of a mobile robot. Benefits of combining game theory and geometry; Analysis of sequence in motion decisions; Characteristics of motion strategies.

Examines the problem in the navigation of a mobile robot in an indoor environment. Development of a navigation method in an unknown environment based on the combination of elementary behaviors; Use of ser-order Takagi-Sugeno fuzzy inference systems; Types of obstacle avoidance behaviors; Robustness of fuzzy logic approach.

Describes the design and implementation on programmable hardware (FPGA) of an algorithm for the detection of self-mobile objects as seen from a mobile robot. Use of foveal images obtained with a special camera whose optical axis points towards the direction of advance; Simplification of the original formulation by the use of log-polar geometry; Differential nature of the approach.

Presents an automatic landmark selection algorithm that allows a mobile robot to select conspicuous landmarks from a continuous stream of sensory perceptions. Absence of pre-installed knowledge or human intervention during the selection process; Use of the algorithm to make mapping mechanisms more efficient and reliable.

Discusses the advantages of incorporating cognition and remembrance capabilities in a sensor-based real-time navigation algorithm of a mobile robot. Features of the algorithm; Enhancement of the robot's performance; Environment of the robot.

Presents a robust position-tracking method for a mobile robot with sonar sensors using Hough transform and probability grids. Computation of a two-dimensional feature space; Combination of the detection of landmark features and matching for sets of features with probability grid updates; Failure of the Markov localisation method.

Presents a novel distributed control scheme of multiple robotic vehicles. Characteristics and capabilities of multi-agent robotic system; Advantage of distributing the formation vectors among mobile robots; Control strategy for each robot.

Proposes a solution for detecting the target and measuring the position and velocity of a mobile robot. Usage of virtual ultrasonic image; Accumulation of the returned ultrasonic signal along the time axis; Representation of the relative speed of the object.

Discusses an action selection mechanism for robots in a robot soccer game using the modular Q-learning system. Learning modules composing the Q-learning architecture; Use of uni-vector field for robot navigation; Effectiveness and applicability of modular Q-learning and uni-vector in a robot soccer experiment.

Explores the integration of basic reflex actions and logical decisions to copy the process of obstacle avoidance behavior of human locomotion. Navigation of mobile robots through unknown environments by using free-target-approach behavior and single side reflex action; Fuzzy-logic based implementation; Use of Perceptional symmetry; Simulation results.

Presents a method of map building for mobile robots, which uses geometric models of the constraints from the environment. Advantages of geometric maps over grid-based maps; Process of constraint identification; Use of geometric objects for constraint modelling; Operations which integrate new sensor readings into existing models; Simulation results.

Discusses the tracking of a sensor's motion which is attached to a person or walking robot. Parameter extraction of a walk motion model through the use of Gabor wavelets and Extended Kalman filters; Butterworth filtering of the motion sensors; Accuracy of the motion tracking system for the obstacle detection process required by a walking person.

Presents a mail distributing service robot, called Mobile Post Distribution System (MOPS). Description of the navigation and control architecture, the position controller and the localization method; Interaction of the robot with its environment; Operating experiences, lessons learned and consequences of using MOPS.

Presents an approach for mobile robot localization using geometric features from a 360 degree laser finder and a monocular vision system. Practicality of the approach under continuous localization during motion in real time; Extraction of the features using physically well-grounded models; Estimation of the fusion and position using an extended Kalman filter.

Presents the theoretical analysis of different methods, which can be used in online path generation for mobile robots with nonholonomic constraints in a partially known workspace. Discussion of different heuristic cost functions and a statement of this function; Simulation tests and experimental results of online control for car-like robot.

Proposes a modular, behavior-based control architecture for autonomous mobile robots, which is suited for 'Learning from Demonstration' experiments in the spatial domain. Online learning of sensory-motor behaviors; Discussion of results using radial basis function-approximation, growing neutral cell structures and probabilistic models for progress estimation.

Presents a unified approach to global localization and position tracking of mobile robot based on a topological map augmented with metric information. Combination of the method with a scan matching technique for tracking multiple location hypotheses over time; Validation of the method with experiments in a series of real world experiments.

Presents a vision system for object verification and localization specially adapted to the needs of mobile robots. Analysis of the object model based on a collection of local features; Interpretation of the tolerance to scale changes; Learned representation of the object model.

Discusses mobile robots with ultrasonic sensors which are useful when navigating environments that contain plants. Difficulty in plant sensing; Sensing capability of continuously transmitted frequency modulated ultrasonic sensor; Improvement of correlation using echo tracking.

Presents a probabilistic technique for estimating the state of dynamic objects in the environment of a mobile service robot. Incorporation of the robot's uncertainty in its position into the state estimation process; Implementation of the method on a real robot; Examples to illustrate the efficiency and robustness of the method.

Proposes the visual attentional selection method in mobile robots based on the amplification of selected stimulus. Relationship between visual attentional computation and ambient light intensity; Parameters for the inhibition and diffusion of retina cell activation; Association of visual stimulus with electrical energy.

Focuses on self-localization system based on omnidirectional color images for autonomous mobile robots. Details on omnidirectional image sensors; Role of optical pre-processing for the learned environment; Association of perceived image with the learning subsystem.

Focuses on the reactive component of a motion control architecture for car-like vehicles. Role of execution monitor (EM) for mobile robots; Details on the fuzzy control mechanisms; Impact of EM on real vehicles.

Deals with location identification of mobile robots equipped with sensors. Utilization of fuzzy triangulation in identifying position and orientation of the robot; Details of the initial and detailed localization procedures; Account of several studies utilizing fuzzy sets and logic in sonar based map building, localization and navigation.

Presents a solution to the problem of simultaneously tracking and extracting a piecewise-linear range representation of a mobile robot's local environment. Specification of the classical framework of the extended Kalman filter; Utilization of an algorithm which is immune to active sensor reorientation; Description of the initialization of feature tracks.

Proposes an innovative measurement scheme using ultrasonic sensors to determine the location and the type of target surface in the working environment of a mobile robot. Classification of different target surfaces; Relationship between the recognition rates of surface types and different distances; Performance of the ultrasonic sensors beyond 15 degrees of angulation.

Examines the problem of mobile robot motion control in partially unknown environments using the sliding-mode fuzzy-logic controller (SMFLC). Development of the hybrid control scheme; Context of the control law; Capability of the reduced complexity SMFLC in controlling the position of the mobile robot.

Presents a system for corridor following in robots based on properties of the human visual system. Use of an interest operator to compute sparse optical flow induced by the translatory motion of the robot; Strategy to achieve the control of the robot; Use of vestibular apparatus to determine the ego-motion of the head; Control of a mobile robot using the foveated vision.

Reports recent results in path planning using an efficient data structure and optimal algorithm. Mobile robots in outdoor unstructured environments; Difference in performance between robots starting with no information regarding the environment versus robots with partial knowledge of the environment.

Proposes a neural network method for real-time motion planning of a mobile robot. Demonstration of effectiveness and efficiency through simulation; Dynamic neural activity landscape of the neural network; Planned real-time comfortable path of motion.

Focuses on a study which examined a method for estimating the position and orientation, or pose, of a mobile robot by registering a segment-based local map observed from the used robot pose and a global map. Overview of related work; Discussion on map alignment and map overlap; Experimental results.

Provides information on a study on the foundations needed in designing an interface system for mobile robot teleoperation. Ecological interface; System architecture; Details on the experiment; Discussion and conclusions.

Provides information on the Robot Sheepdog Project's mobile robot that gathers a flock of ducks and maneuvers it to a goal position. Features of the robot vehicle; Experimental results on the mobile robot; Scope and limitations of the experiments.

Addresses the problem of visual-based navigation of a mobile robot in indoors environments. Description of the geometric foundations related to the vanishing point determination and the servoing strategy; How appearance-based methods can be used for extending the autonomy achieved by servoing methods alone; Conclusion.

Presents a strategy for the navigation of mobile robots in open sparse uncertain environments. Functionalities required to implement the strategy; Implementation of the strategy; Effectiveness of the approach.

Presents a technique for on-line extraction of natural point landmarks in an indoor environment of mobile robots. Description of basic algorithm for natural point landmark extraction; Experiments on localization ability; Navigation experiments on the Nomad 200 robot.

Presents a two-layer mobile robot localization solution based on a Laser Range Scanner applied on reconstructed three-dimensional models. Description of the RESOLV system project; Details on the frame and approximate localization algorithm; Conclusion.

Presents information on a study which proposed a virtual target approach for resolving the limit cycle problem in navigation of a behavior-based mobile robot. Analysis of the cause of the limit cycle; Description of the proposed virtual target approach and its algorithm; Simulation results; Conclusions.

This work considers the problem of chaotic path planning, using an improved memory technique to boost performance. In this application, the dynamics of two simple chaotic maps are first used to generate a pseudo-random bit generator. Using this as a source, a series of navigation commands are generated and used by an autonomous robot to explore an area, while maintaining a random and unpredictable motion. This navigation strategy can bring overall area coverage, but also yields numerous revisits to previous cells. Here, a memory technique is applied to limit the chaotic motion of the robot to adjacent cells with the least number of visits, leading to overall improvement in performance. Numerical simulations are performed to evaluate the path planning strategy. The simulation results showcase a major improvement in coverage performance compared to the memory-free technique and also compared to an inverse pheromone technique previously developed by the authors. Also, the number of multiple visits to previous cells is significantly reduced with the proposed technique. • Creation of a Bit Generator for producing the motion commands of a mobile robot. • Randomness validation of the produced motion sequence by using the NIST statistical package. • Coverage rate computation for a given terrain. • Insertion of the Memory Technique for improving the results. • Comparison between memory-free and Memory Technique. [ABSTRACT FROM AUTHOR]

In the last decades, visual target tracking has been one of the primary research interests of the Robotics research community. The recent advances in Deep Learning technologies have made the exploitation of visual tracking approaches effective and possible in a wide variety of applications, ranging from automotive to surveillance and human assistance. However, the majority of the existing works focus exclusively on passive visual tracking, i.e. , tracking elements in sequences of images by assuming that no actions can be taken to adapt the camera position to the motion of the tracked entity. On the contrary, in this work, we address visual active tracking, in which the tracker has to actively search for and track a specified target. Current State-of-the-Art approaches use Deep Reinforcement Learning (DRL) techniques to address the problem in an end-to-end manner. However, two main problems arise: (i) most of the contributions focus only on discrete action spaces, and the ones that consider continuous control do not achieve the same level of performance; and (ii) if not properly tuned, DRL models can be challenging to train, resulting in considerably slow learning progress and poor final performance. To address these challenges, we propose a novel DRL-based visual active tracking system that provides continuous action policies. To accelerate training and improve the overall performance, we introduce additional objective functions and a Heuristic Trajectory Generator (HTG) to facilitate learning. Through extensive experimentation, we show that our method can reach and surpass other State-of-the-Art approaches performances, and demonstrate that, even if trained exclusively in simulation, it can successfully perform visual active tracking even in real scenarios. • A novel Deep Reinforcement Learning approach for Visual Active Tracking is proposed. • The method is able to compute action policies on the continuous space. • A novel Heuristic Trajectories Generator algorithm is used to improve learning. • Training is performed only in simulated environments. • The approach is evaluated in a real-world scenario with no fine-tuning. [ABSTRACT FROM AUTHOR]

Snake robots have advantages of terrain adaptability over wheeled mobile robots and traditional articulated robot arms because of their limbless thin body structure and high flexibility. They have extensive applications in tasks such as rescue, disaster recovery, inspection and minimally invasive surgery. Current research on snake robots is mainly focused on snake-like locomotion and the embodiment of these motion gaits for different applications. Modular structure and real-time control algorithms are two key aspects for snake robots operating in constrained environments. This review will attempt to address both. First, a review on the snake motion and the body structure is provided, which outlines the biological foundation of all snake robots. This is followed by the mechanical structure of snake robots, especially the structure of elemental snake modules. Finally, control algorithms for variant terrain contours and obstacle avoidance are discussed. The review also outlines emerging application areas and potential future directions of snake robots. • Structure and control of snake robots in constrained environments are addressed. • A review on the locomotion and body structure of snakes is provided. • Mechanical structure and actuators of snake robots are introduced. • Control for variant terrain contours and obstacle avoidance are discussed. • The review outlines emerging application areas and potential future directions. [ABSTRACT FROM AUTHOR]