Search

We study the problem of searching for and tracking a collection of moving targets using a robot with a limited field-of-view (FOV) sensor. The actual number of targets present in the environment is not known a priori. We propose a search and tracking framework based on the concept of Bayesian random finite sets (RFSs). Specifically, we generalize the Gaussian mixture probability hypothesis density (GM-PHD) filter which was previously applied for tracking problems to allow for simultaneous search and tracking with a limited FOV sensor. The proposed framework can extract individual target tracks as well as estimate the number and the spatial density of targets. We also show how to use the Gaussian process (GP) regression to extract and predict unknown target trajectories in this framework. We demonstrate the efficacy of our techniques through representative simulations and a real data collected from an aerial robot. Note to Practitioners—This article is motivated by search-and-rescue operations where a robot with limited field-of-view (FOV) is used to search and track lost targets. This article presents an estimation and planning framework to estimate the position of targets and track them over time. The key feature of the proposed algorithm is that it can deal with an unknown and varying number of targets. The framework can also deal with an unknown motion model for targets which itself can be complex. The algorithm is shown to be robust to a poor initialization and can handle an initial belief which overestimates or underestimates the actual number of targets. The proposed scheme includes various user-defined parameters. It is recommended to tune these parameters a priori using simulations for a better performance. Incorporating a multirobot approach into the proposed algorithm and finding a better planning strategy that minimizes the time are potential future works. [ABSTRACT FROM AUTHOR]

In this article, we present the design and experimental validation of source-seeking control algorithms for a unicycle mobile robot that is equipped with novel 3-D-printed flexible graphene-based piezoresistive airflow sensors. Based solely on a local gradient measurement from the airflow sensors, we propose and analyze a projected gradient ascent algorithm to solve the source-seeking problem. In the case of partial sensor failure, we propose a combination of extremum-seeking control with our projected gradient ascent algorithm. For both control laws, we prove the asymptotic convergence of the robot to the source. Numerical simulations were performed to validate the algorithms, and experimental validations are presented to demonstrate the efficacy of the proposed methods. [ABSTRACT FROM AUTHOR]

This article proposes indirect/direct learning control schemes for wireless sensor and mobile robot networks to cover an environment according to the density function, which is the distribution of an important quantity within the environment. When stationary sensors cooperate with mobile robots, the density estimation can be enhanced by using nonstationary basis functions to relax the assumption of matching conditions in the previous approach. To improve the density function estimation, this study employs an expectation–maximization algorithm and log-likelihood, which maximizes the similarity between the proposed normalized density and normalized coverage function. Subsequently, the adaptive weighting algorithm is combined with the proposed indirect coverage control for tunable basis centers and the weighting of the basis functions. For direct coverage control, mobile robots are driven to cover the regions of higher importance while simultaneously estimating the density function utilizing a sensory model function. We prove that the Lloyd algorithm is a special case of the direct method when the density function and Voronoi partitions are available. The efficiency of the proposed methods is confirmed in numerical examples and semiexperiments. [ABSTRACT FROM AUTHOR]

Robots in swarms take advantage of localization infrastructure, such as a motion capture system or global positioning system (GPS) sensors to obtain their global position, which can then be communicated to other robots for swarm coordination. However, the availability of localization infrastructure needs not to be guaranteed, e.g., in GPS-denied environments. Likewise, the communication overhead associated with broadcasting locations may be undesirable. For reliable and versatile operation in a swarm, robots must sense each other and interact locally. Motivated by this requirement, we propose an onboard relative localization framework for multirobot systems. The setup consists of an anchor robot with three onboard ultrawideband (UWB) sensors and a tag robot with a single onboard UWB sensor. The anchor robot utilizes the three UWB sensors to estimate the tag robot’s location by using its onboard sensing and computational capabilities solely, without explicit interrobot communication. Because the anchor UWB sensors lack the physical separation that is typical in fixed UWB localization systems, we introduce filtering methods to improve the estimation of the tag’s location. In particular, we utilize a mixture Monte Carlo localization (MCL) approach to capture maneuvers of the tag robot with acceptable precision. We validate the effectiveness of our algorithm with simulations as well as indoor and outdoor field experiments on a two-drone setup. The proposed mixture MCL algorithm yields highly accurate estimates for various speed profiles of the tag robot and demonstrates superior performance over the standard particle filter and the extended Kalman filter. [ABSTRACT FROM AUTHOR]

The range-only simultaneous localization and mapping (RO-SLAM) problem is considered in this article. This is a difficult problem since bearing measurements are not available. Also, the frequent outliers induced by the sensor nature (typically sonar or radio pulses) render the RO-SLAM problem more challenging. The non-Gaussian distribution found in the RO-SLAM problem makes that the landmarks cannot be directly updated using the approach based on the Gaussian assumption. We propose a region-based particle filter and a proper likelihood model to realize the reduction of particle numbers and allow the introduction of new landmarks without delayed processing. A transition mode is designed from the loosely coupled way to the tightly coupled way and used to estimate the pose of the robot and locations of landmarks. Furthermore, two detection methods are, respectively, used to remove the outliers generated by the range-only sensors according to the distribution status of the landmark estimation. The proposed algorithm is evaluated through simulations and experiments with a range-only sensor, i.e., Ultrawideband (UWB). Furthermore, the obtained results are compared with the classical algorithm. The simulation and experiment results verify the validity and superiority of the proposed RO-SLAM algorithm. [ABSTRACT FROM AUTHOR]

Building an efficient and reliable collision perception visual system is a challenging problem for future robots and autonomous vehicles. The biological visual neural networks, which have evolved over millions of years in nature and are working perfectly in the real world, could be ideal models for designing artificial vision systems. In the locust’s visual pathways, a lobula giant movement detector (LGMD), that is, the LGMD2, has been identified as a looming perception neuron that responds most strongly to darker approaching objects relative to their backgrounds; similar situations which many ground vehicles and robots are often faced with. However, little has been done on modeling the LGMD2 and investigating its potential in robotics and vehicles. In this article, we build an LGMD2 visual neural network which possesses the similar collision selectivity of an LGMD2 neuron in locust via the modeling of biased-ON and -OFF pathways splitting visual signals into parallel ON/OFF channels. With stronger inhibition (bias) in the ON pathway, this model responds selectively to darker looming objects. The proposed model has been tested systematically with a range of stimuli including real-world scenarios. It has also been implemented in a micro-mobile robot and tested with real-time experiments. The experimental results have verified the effectiveness and robustness of the proposed model for detecting darker looming objects against various dynamic and cluttered backgrounds. [ABSTRACT FROM AUTHOR]

This article proposes a robust visual following approach with a deep learning-based person detector, a Kalman filter (KF), and a reidentification module. The KF is introduced to predict the position of the target person, and its state is updated by the associated detection result. To deal with severe distractions and even full occlusion, the reidentification module with an identification model, a verification model, and an appearance gallery is employed in multi-person disturbing environments. Without any customized markers, the proposed approach can follow the target person steadily, and it is robust to occlusion and posture changes of the target person. Experiments results validate the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

A simple hardware array for obtaining highresolution structural vibration data will enable future civil infrastructures to become general sensing platforms that are easy and practical to deploy and maintain over the lifetime of a building. [ABSTRACT FROM AUTHOR]

A fundamental challenge of mobile sensor networks is automated active reconfiguration of sensors in response to environmental stimuli in order to maximize their total sensing quality (or minimize their total sensing cost) of events occurring over an environment. In this paper, given an event distribution over a convex environment, we consider mobile isotropic sensors with adjustable sensing range and propose a new family of provably correct reactive coverage control algorithms for both continuous- and discrete-time sensor dynamics. The proposed coverage control algorithms constantly (re)configure sensor positions and sensing ranges in order to minimize a statistical distance, in particular, an $f$ -divergence, between the event distribution over the environment and the overall event detection probability of sensors. We show that the standard Voronoi-based coverage control law of homogeneous mobile sensor networks is a special case of our framework where the event detection probability of each sensor has a Gaussian form, the statistical distance is set to be the Kullback–Leibler (KL) divergence and sensor allocation is performed based on Voronoi diagrams. To increase the practicality of our framework, we also present its integration with a Voronoi-based collision avoidance strategy for disk-shaped sensor bodies and its extension to differential drive sensor dynamics, while retaining the stability properties. [ABSTRACT FROM AUTHOR]

This paper describes a novel probabilistic sensor fusion framework aimed at improving obstacle detection accuracy and classification of various targets experienced in the Maritime RobotX Challenge. In both the 2014 and 2016 Maritime RobotX Challenges, it was found that detecting obstacles using LIDAR only and classifying obstacles using vision only can be challenged by environmental conditions, such as glare from the sun or by objects such as the spherical black buoys from the obstacle field that disperses LIDAR rays. In this paper, a new multimodal sensor fusion approach is proposed that combines data streams from perception sensors, such as LIDAR, RADAR, and cameras, to improve the robustness of detection and classification performance over a single sensor method. Using data collected from both the 2014 and 2016 Maritime RobotX Challenges, an evaluation of the perception framework is provided. The proposed detection and classification framework is now being transferred to the queensland university of technology (QUT) autonomous surface vehicle to improve overall mapping accuracy and task execution. [ABSTRACT FROM AUTHOR]

A path planning and control approach of a non-holonomic three-wheeled mobile robot (WMR) for online navigation in road following and roundabout environments is presented in this paper. We proposed a complete navigation algorithm that enables the WMR to autonomously navigate on the road with various scenarios. With such an algorithm, the robot is able to localize itself within the road environment and find a collision free-path starting from a pre-defined start position to a goal point using a novel approach called laser simulator (LS). The path planning and roundabout detection are determined based on LS and sensor fusion of a laser range finder, camera, and odometry measurements. The sensor fusion algorithm is used to remove noises and uncertainties from sensors’ data and provide optimum measurements for path planning. A robot motion control scheme is used for the purpose of controlling the kinematic parameters of WMR using a resolved acceleration control coupled with an active force control for rejecting the disturbances. Experimental results show the capability of the proposed algorithms to robustly drive the robot on the road following and roundabout environments. [ABSTRACT FROM AUTHOR]

This paper deals with the bearing angle error (BAE) estimation problem associated with a one-axis gimballed ultrasonic seeker, which can be used as a low-cost sensor for mobile robot applications. Under the assumptions that the target has an ultrasonic transmitter and the gimbal control loops work properly, the moving target tracking reduces to an estimation problem of a piecewise constant BAE using real sinusoids measured by a receiver array mounted on the gimbal. To effectively cope with the real-time implementation issue, our estimation problem is formulated using linear robust state estimation and is resolved using the constrained robust weighted least squares method. Through experimental results obtained from a hardware-in-the-loop simulation (HILS) approach, it is demonstrated that the proposed scheme provides more reliable and consistent estimation performance than existing approaches. [ABSTRACT FROM PUBLISHER]

Flying in unknown environments may lead to unforeseen collisions, which may cause serious damage to the robot and/or its environment. In this context, fast and robust collision detection combined with safe reaction is, therefore, essential and may be achieved using external wrench information. Also, deliberate physical interaction requires a control loop designed for such a purpose and may require knowledge of the contact wrench. In principle, the external wrench may be measured or estimated. Whereas measurement poses large demands on sensor equipment, additional weight, and overall system robustness, in this paper we present a novel model-based method for external wrench estimation in flying robots. The algorithm is based on the onboard inertial measurement unit and the robot's dynamics model only. We design admittance and impedance controllers that use this estimate for sensitive and robust physical interaction. Furthermore, the performance of several collision detection and reaction schemes is investigated in order to ensure collision safety. The identified collision location and associated normal vector located on the robot's convex hull may then be used for sensorless tactile sensing. Finally, a low-level collision reflex layer is provided for flying robots when obstacle avoidance fails, also under wind influence. Our experimental and simulation results show evidence that the methodologies are easily implemented and effective in practice. [ABSTRACT FROM PUBLISHER]

We propose a lightweight and fast learning algorithm for classifying the features of an unknown terrain that a robot is navigating in. Most of the existing research on unknown terrain classification by mobile robots relies on a single powerful classifier to correctly identify the terrain using sensor data from a single sensor like laser or camera. In contrast, our proposed approach uses multiple modalities of sensed data and multiple, weak but less-complex classifiers for classifying the terrain types. The classifiers are combined using an ensemble learning algorithm to improve the algorithm’s training rate as compared to an individual classifier. Our algorithm was tested with data collected by navigating a four-wheeled, autonomous robot, called Explorer, over different terrains including brick, grass, rock, sand, and concrete. Our results show that our proposed approach performs better with up to 63% better prediction accuracy for some terrains as compared to a support vector machine (SVM)-based learning technique that uses sensor data from a single sensor. Despite using multiple classifiers, our algorithm takes only a fraction (1/65) of the time on average, as compared to the SVM technique. [ABSTRACT FROM AUTHOR]

In this article, a localization system for a mobile robot, using a top-down multisensors approach and a map of the environment, is proposed. Popular methods try to optimize a global cost, track multihypothesis, or reduce the problem by using multisensors. These approaches are bottom-up: Each sensor data is analyzed even if it is not relevant [like a global positioning system (GPS) in an indoor environment]. [ABSTRACT FROM AUTHOR]

This paper proposes a robust distance-based tracking control method for two-wheeled differential drive mobile robots (WMRs) in the presence of kinematic disturbances, where only the visual image information between the WMR and the target is used to obtain distances, and the velocity sensors are not available. To achieve the tracking control of the WMR toward the target, a vision system is used so as to obtain the relative posture information between the WMR and the target. In particular, the WMR is assumed to move without velocity sensors, and the posture and velocity information for the WMR and that of the target is unavailable. The necessary information of the maximum velocities between the WMR and the target is thus estimated in an online manner. Correspondingly, a robust backstepping-like feedback linearization method is proposed to simultaneously compensate for the unknown velocities and kinematic disturbances. Experiments have been performed for the WMR with the vision system to verify that the proposed method can allow the WMR to track the target in the presence of kinematic disturbances, even without using the velocity sensors. [ABSTRACT FROM PUBLISHER]

On 18 November 2014, a team of four autonomous CoBot robots reached 1,000-km of overall autonomous navigation, as a result of a 1,000-km challenge that the authors had set three years earlier. The authors are frequently asked for the lessons learned, as well as the performance results. In this article, they introduce the challenge and contribute a detailed presentation of technical insights as well as quantitative and qualitative results. They have previously presented the algorithms for the individual technical contributions, namely robot localization, symbiotic robot autonomy, and robot task scheduling. In this article, they present the data collected over the 1,000-km challenge and analyze it to evaluate the accuracy and robustness of the localization algorithms on the CoBots. Furthermore, they present technical insights into the algorithms, which they believe are responsible for the robots' continuous robust performance. [ABSTRACT FROM PUBLISHER]

In this paper, we present a novel vision-based detection system able to extend the detection range of vehicles or mobile robots. The proposed system is used to detect and track moving targets from near-to-far ranges and covers a wide range of more than 130 m without decreasing detection accuracy. Typical examples of targets include traffic signs, vehicles, animals, and pedestrians. In this paper, the detection and tracking of moving pedestrians from near-to-far ranges is investigated. The proposed system is composed of two identical cameras. The first camera is equipped with a short focal length lens to detect and track pedestrians in near-to-mid range, and the second camera with a long focal length lens is used to detect and track pedestrians in mid-to-far range. To synchronize the detection results of both cameras and to eliminate repeated measurements, two synchronization algorithms were developed. The tracking process is applied after the detection, and it is used to track and predict the future motion and direction of pedestrian. To prevent vehicle-target collisions, two algorithms that generate alert and danger warnings are developed. A mathematical model based on the fundamental physics of the camera and lens is developed to illustrate the feasibility of our work. Finally, we conducted many experiments in large open-air parking lots and on Ottawa roads to show the applicability of our system. [ABSTRACT FROM AUTHOR]

This brief presents a control system for a snake robot based on range sensor data that semiautonomously aids the robot in avoiding collisions with obstacles. In the proposed system, an operator indicates the desired velocity of the first link of the robot using a joystick, and the joint input which accomplishes both the desired velocity of the first link and collision avoidance between subsequent links and obstacles is automatically calculated by the controller, which selects the links needed to be grounded and exploits redundancy. The controller uses real-time data from range sensors for obstacle positions. The experimental system, which has range sensors and the function generating environmental map using simultaneous localization and mapping, was developed with decreasing calculation cost, and experiments were performed to verify the effectiveness of the proposed system in unknown environments. [ABSTRACT FROM PUBLISHER]