Search

In this article, we present a new controller for stable and robust walking control of ATRIAS, an underactuated bipedal robot designed based on the spring-loaded inverted pendulum (SLIP) model. We propose a forced-oscillation scheme for control of vertical motion, which we prove to be stable and contractive. Moreover, we prove that, through some mild assumptions, the dynamics of the system can be written in a hierarchical form that decouples the stability analyses of the horizontal and vertical directions. We leverage these properties to find a stabilizing class of functions for foot placement. The torso control is also proved to be decoupled using singular perturbation theory and is stabilized through a feedback linearization controller. We also take advantage of the proposed framework's flexibility and extend it to include a new reflex-based uneven-terrain walking control scheme. We test the controller for various desired walking speeds (0 to 2.5 m/s), for stepping up and down unexpected obstacles (15 cm), and for high-speed walking on a random uneven terrain (up to 10 cm of step-ups and step-downs and up to 1.8 m/s). The results show successful performance of the controller and its stability and robustness against various perturbations. [ABSTRACT FROM AUTHOR]

One of the most difficult challenges in robot motion planning is to account for the behavior of other moving agents, such as humans. Commonly, practitioners employ predictive models to reason about where other agents are going to move. Though there has been much recent work in building predictive models, no model is ever perfect: an agent can always move unexpectedly, in a way that is not predicted or not assigned sufficient probability. In such cases, the robot may plan trajectories that appear safe but, in fact, lead to collision. Rather than trust a model's predictions blindly, we propose that the robot should use the model's current predictive accuracy to inform the degree of confidence in its future predictions. This model confidence inference allows us to generate probabilistic motion predictions that exploit modeled structure when the structure successfully explains human motion, and degrade gracefully whenever the human moves unexpectedly. We accomplish this by maintaining a Bayesian belief over a single parameter that governs the variance of our human motion model. We couple this prediction algorithm with a recently proposed robust motion planner and controller to guide the construction of robot trajectories that are, to a good approximation, collision-free with a high, user-specified probability. We provide extensive analysis of the combined approach and its overall safety properties by establishing a connection to reachability analysis, and conclude with a hardware demonstration in which a small quadcopter operates safely in the same space as a human pedestrian. [ABSTRACT FROM AUTHOR]

This paper defines the research area of Diversity-enhanced Autonomy in Robot Teams (DART), a novel paradigm for the creation and design of policies for multi-robot coordination. Although current approaches to multi-robot coordination have been successful in structured, well-understood environments, they have not been successful in unstructured, uncertain environments, such as disaster response. Although robot hardware has advanced significantly in the past decade, the way we solve multi-robot problems has not. Even with significant advances in the field of multi-robot systems, the same problem-solving paradigm has remained: assumptions are made to simplify the problem, and a solution is optimized for those assumptions and deployed to the entire team. This results in brittle solutions that prove incapable if the original assumptions are invalidated. This paper introduces a new multi-robot problem-solving paradigm which uses a diverse set of control policies that work together synergistically within the same team of robots. Such an approach will make multi-robot systems more robust in unstructured and uncertain environments, such as in disaster response, environmental monitoring, and military applications, and allow multi-robot systems to extend beyond the highly structured and highly controlled environments where they are successful today. [ABSTRACT FROM AUTHOR]

Collaborative object transportation using multiple MAV with limited communication is a challenging problem. In this paper, we address the problem of multiple MAV mechanically coupled to a bulky object for transportation purposes without explicit communication between agents. The apparent physical properties of each agent are reshaped to achieve robustly stable transportation. Parametric uncertainties and unmodeled dynamics of each agent are quantified and techniques from robust control theory are employed to choose the physical parameters of each agent to guarantee stability. Extensive simulation analysis and experimental results show that the proposed method guarantees stability in worst-case scenarios. [ABSTRACT FROM AUTHOR]

One of the major challenges for mobile robots in human-shaped environments is navigating stairways. This study presents a method for accurately detecting, localizing, and estimating the characteristics of stairways using point cloud data. The main challenge is the wide variety of different structures and shapes of stairways. This challenge is often aggravated by an unfavorable position of the sensor, which leaves large parts of the stairway occluded. This can be further aggravated by sparse point data. We overcome these difficulties by introducing a three-dimensional graph-based stairway-detection method combined with competing initializations. The stairway graph characterizes the general structural design of stairways in a generic way that can be used to describe a large variety of different stairways. By using multiple ways to initialize the graph, we can robustly detect stairways even if parts of the stairway are occluded. Furthermore, by letting the initializations compete against each other, we find the best initialization that accurately describes the measured stairway. The detection algorithm utilizes a plane-based approach. We also investigate different planar segmentation algorithms and experimentally compare them in an application-orientated manner. Our system accurately detects and estimates the stairway parameters with an average error of only 2 . 5 mm for a variety of stairways including ascending, descending, and spiral stairways. Our method works robustly with different depth sensors for either small- or large-scale environments and for dense and sparse point cloud data. Despite this generality, our system's accuracy is higher than most state-of-the-art stairway-detection methods. [ABSTRACT FROM AUTHOR]

Grasping in constrained environments is, to this day, an ongoing research topic. Objects can rarely be grasped from arbitrary directions, hence the need to study the options available to grasp them. This paper proposes a gripper capable of grasping small or thin objects that cannot be directly pinch-grasped. The focus is placed on objects that lie on hard surfaces. The proposed approach uses a quasistatic method referred to as scooping while implementing a passive thumb to compensate for manipulator positioning errors. Hence, the robot arm does not need to be moved while the gripper is grasping an object, similarly to a human hand performing a precision grasp. The design approach is presented and the main design choice, namely the use of epicyclic gear trains instead of conventional revolute joints, is explained. The implementation of the proposed approach to the gripper design is shown. We explain how parallel pinch grasps and large grasping forces are achieved even though the mechanism does not follow the usual parallelogram four-bar implementation of parallel pinch mechanisms. The experimental validation of the proposed concept is then presented by picking up a set of test objects in sequence and demonstrating some variants of the method that expand on the concept. [ABSTRACT FROM AUTHOR]

Terrain classification is a critical component of any autonomous mobile robot system operating in unknown real-world environments. Over the years, several proprioceptive terrain classification techniques have been introduced to increase robustness or act as a fallback for traditional vision based approaches. However, they lack widespread adaptation due to various factors that include inadequate accuracy, robustness and slow run-times. In this paper, we use vehicle-terrain interaction sounds as a proprioceptive modality and propose a deep long-short term memory based recurrent model that captures both the spatial and temporal dynamics of such a problem, thereby overcoming these past limitations. Our model consists of a new convolution neural network architecture that learns deep spatial features, complemented with long-short term memory units that learn complex temporal dynamics. Experiments on two extensive datasets collected with different microphones on various indoor and outdoor terrains demonstrate state-of-the-art performance compared to existing techniques. We additionally evaluate the performance in adverse acoustic conditions with high-ambient noise and propose a noise-aware training scheme that enables learning of more generalizable models that are essential for robust real-world deployments. [ABSTRACT FROM AUTHOR]

The problem of simultaneous localisation and mapping (SLAM) has been addressed in numerous ways with different approaches aiming to produce faster, more robust solutions that yield consistent maps. This focus, however, has resulted in a number of solutions that perform poorly in challenging real life scenarios. In order to achieve improved performance and map quality this article proposes a novel method to construct informative Bayesian mapping priors through a multi-objective optimisation of prior map design variables defined using a source of prior information. This concept is explored for 2D occupancy grid SLAM, constructing such priors by extracting structural information from architectural drawings and identifying optimised prior values to assign to detected walls and empty space. Using the proposed method a contextual optimised prior can be constructed. This prior is found to yield better quantitative and qualitative performance than the commonly used non-informative prior, yielding an increase of over 20% in the F 2 metric. This is achieved without adding to the computational complexity of the SLAM algorithm, making it a good fit for time critical real life applications such as search and rescue missions. [ABSTRACT FROM AUTHOR]

This paper focuses on devising a control policy that is inspired by human strategy to enable robots to perform surface–surface contact between a hand-held object (e.g. a box) and the environment (e.g. table). We assume the object’s shape is partially known and consider uncertainties in both the environment and the object grasp. Our analysis on ten untrained subjects indicates that during this task: (1) the subjects start decreasing the angular velocity before the complete alignment (in contrast to existing approaches), and (2) they do not control the contact force to remain at a fixed value. Our study is also consistent with the hypothesis that the subjects determine an over-estimate of the relative angle between the object in hand and the environment. Based on these observations we propose a novel control policy, called Surface-Surface Contact Primitive (SSCP), which can perform the surface–surface alignment task with only partial information about the hand-held object and the environment. Furthermore, SSCP only requires a rough estimate of the surface normal vector and does not rely on either the estimation of contact type (i.e. point or edge contact) or locations of contact points. We evaluate the performance of the proposed controller on a set of robot experiments using two seven-degree-of-freedom robots, one for imposing uncertainty to the environment, and the other to perform the experiment. We show the applicability of the controller on four objects with different geometries, its generalization to four different surface materials, and its robustness to uncertainty in environment and grasp as well as external perturbations. [ABSTRACT FROM AUTHOR]

We consider the problem of computing a minimum uncertainty path for an autonomous vehicle from a start to a destination location in the presence intermittent sensing, modeled as a stochastic process, in addition to process and measurement noise. We introduce the use of a novel bound on the maximum eigenvalue of the estimation error covariance matrix as the cost function for belief space planning. Our main contributions are three-fold. We first derive an analytic bound on the performance of a state estimator under sensor misdetection (intermittency) occurring stochastically over time. Second, we use this bound as a proxy for the expected maximum eigenvalue evolution in a sample-based path planning algorithm to produce a path that trades off accuracy and robustness. This extends the recent body of work on planning under uncertainty to include the fact that sensors may not provide any measurement owing to misdetection. Computational results demonstrate the benefit of the approach and comparisons are made with the state of the art in path planning in belief space. Third, and finally, we establish theoretically that the proposed algorithm possesses the optimal substructure property, i.e. the algorithm returns an optimal path relative to the bound treated as a proxy for the expected maximum eigenvalue evolution. [ABSTRACT FROM AUTHOR]

This paper presents a Robust Constrained Learning-based Nonlinear Model Predictive Control (RC-LB-NMPC) algorithm for path-tracking in off-road terrain. For mobile robots, constraints may represent solid obstacles or localization limits. As a result, constraint satisfaction is required for safety. Constraint satisfaction is typically guaranteed through the use of accurate, a priori models or robust control. However, accurate models are generally not available for off-road operation. Furthermore, robust controllers are often conservative, since model uncertainty is not updated online. In this work our goal is to use learning to generate low-uncertainty, non-parametric models in situ. Based on these models, the predictive controller computes both linear and angular velocities in real-time, such that the robot drives at or near its capabilities while respecting path and localization constraints. Localization for the controller is provided by an on-board, vision-based mapping and navigation system enabling operation in large-scale, off-road environments. The paper presents experimental results, including over 5 km of travel by a 900 kg skid-steered robot at speeds of up to 2.0 m/s. The result is a robust, learning controller that provides safe, conservative control during initial trials when model uncertainty is high and converges to high-performance, optimal control during later trials when model uncertainty is reduced with experience. [ABSTRACT FROM AUTHOR]

Visual Odometry is a key technology for robust and accurate navigation of unmanned aerial vehicles in a number of low altitude applications (<120 m), particularly in environments where access to a global positioning system is possible, but not guaranteed. Navigating via vision alone reduces dependence on a global positioning system and other global navigation satellite systems, enhancing navigation robustness even in the presence of jamming, spoofing or long dropouts. To date, however, most demonstrations of visual odometry are in close proximity to the ground or other structures and are often implemented as a monocular camera combined with inertial sensing, rather than vision alone, to account for scale drift. Stereo visual odometry has received little attention for applications beyond 30 m altitude due to the generally poor performance of stereo rigs for these extremely small baseline-to-depth ratios, otherwise termed long-range stereo. This paper demonstrates stereo visual pose estimation at altitudes of up to 120 m above ground level on a small fixed-wing unmanned aerial vehicle by adapting the traditional stereo visual odometry paradigm to explicitly account for inaccurate triangulation and poorly observed scale from the stereo baseline. In addition, issues related to long-range stereo such as biased sensing are investigated to justify the approach, and a novel bundle adjustment algorithm is presented capable of handling vibration induced structural deformation between the cameras. This is achieved by continually optimizing the stereo transform within a set of inequality bounds. Results are presented demonstrating the algorithm on field-gathered data from a 2 m wingspan fixed-wing unmanned aerial vehicle flying at 30-120 m altitude over a 6.5 km trajectory. [ABSTRACT FROM AUTHOR]

For the control of unmanned aerial vehicles (UAVs) in GPS-denied environments, cameras have been widely exploited as the main sensory modality for addressing the UAV state estimation problem. However, the use of visual information for ego-motion estimation presents several theoretical and practical difficulties, such as data association, occlusions, and lack of direct metric information when exploiting monocular cameras. In this paper, we address these issues by considering a quadrotor UAV equipped with an onboard monocular camera and an inertial measurement unit (IMU). First, we propose a robust ego-motion estimation algorithm for recovering the UAV scaled linear velocity and angular velocity from optical flow by exploiting the so-called continuous homography constraint in the presence of planar scenes. Then, we address the problem of retrieving the (unknown) metric scale by fusing the visual information with measurements from the onboard IMU. To this end, two different estimation strategies are proposed and critically compared: a first exploiting the classical extended Kalman filter (EKF) formulation, and a second one based on a novel nonlinear estimation framework. The main advantage of the latter scheme lies in the possibility of imposing a desired transient response to the estimation error when the camera moves with a constant acceleration norm with respect to the observed plane. We indeed show that, when compared against the EKF on the same trajectory and sensory data, the nonlinear scheme yields considerably superior performance in terms of convergence rate and predictability of the estimation. The paper is then concluded by an extensive experimental validation, including an onboard closed-loop control of a real quadrotor UAV meant to demonstrate the robustness of our approach in real-world conditions. [ABSTRACT FROM AUTHOR]

We investigate the premise that robust grasping performance is enabled by exploiting constraints present in the environment. These constraints, leveraged through motion in contact, counteract uncertainty in state variables relevant to grasp success. Given this premise, grasping becomes a process of successive exploitation of environmental constraints, until a successful grasp has been established. We present support for this view found through the analysis of human grasp behavior and by showing robust robotic grasping based on constraint-exploiting grasp strategies. Furthermore, we show that it is possible to design robotic hands with inherent capabilities for the exploitation of environmental constraints. [ABSTRACT FROM AUTHOR]

This paper reports on an exact, real-time solution for server–client cooperative localization over a faulty and extremely bandwidth-limited underwater communication channel. Our algorithm, termed the origin state method, enables a ‘server’ vehicle to broadcast its navigation information to multiple ‘client’ vehicles over a bandwidth-limited and faulty communication channel. The server’s broadcasted pose-graph can be used in conjunction with an estimator on the client to exactly reproduce the corresponding server–client centralized estimate. We present an evaluation over an extensive real-time field implementation of the proposed algorithm for a multi-agent autonomous underwater vehicle network using underwater acoustic modems to communicate in a synchronous-clock transmission framework. [ABSTRACT FROM AUTHOR]

This paper summarizes recent activities carried out for the development of an innovative anthropomorphic robotic hand called the DEXMART Hand. The main goal of this research is to face the problems that affect current robotic hands by introducing suitable design solutions aimed at achieving simplification and cost reduction while possibly enhancing robustness and performance. While certain aspects of the DEXMART Hand development have been presented in previous papers, this paper is the first to give a comprehensive description of the final hand version and its use to replicate human-like grasping. In this paper, particular emphasis is placed on the kinematics of the fingers and of the thumb, the wrist architecture, the dimensioning of the actuation system, and the final implementation of the position, force and tactile sensors. The paper focuses also on how these solutions have been integrated into the mechanical structure of this innovative robotic hand to enable precise force and displacement control of the whole system.Another important aspect is the lack of suitable control tools that severely limits the development of robotic hand applications. To address this issue, a new method for the observation of human hand behavior during interaction with common day-to-day objects by means of a 3D computer vision system is presented in this work together with a strategy for mapping human hand postures to the robotic hand. A simple control strategy based on postural synergies has been used to reduce the complexity of the grasp planning problem. As a preliminary evaluation of the DEXMART Hand’s capabilities, this approach has been adopted in this paper to simplify and speed up the transfer of human actions to the robotic hand, showing its effectiveness in reproducing human-like grasping. [ABSTRACT FROM AUTHOR]

Long-term autonomous mobile robot operation requires considering place recognition decisions with great caution. A single incorrect decision that is not detected and reconsidered can corrupt the environment model that the robot is trying to build and maintain. This work describes a consensus-based approach to robust place recognition over time, that takes into account all the available information to detect and remove past incorrect loop closures. The main novelties of our work are: (1) the ability of realizing that, in light of new evidence, an incorrect past loop closing decision has been made; the incorrect information can be removed thus recovering the correct estimation with a novel algorithm; (2) extending our proposal to incremental operation; and (3) handling multi-session, spatially related or unrelated scenarios in a unified manner. We demonstrate our proposal, the RRR algorithm, on different odometry systems, e.g. visual or laser, using different front-end loop-closing techniques. For our experiments we use the efficient graph optimization framework g2o as back-end. We back our claims up with several experiments carried out on real data, in single and multi-session experiments showing better results than those obtained by state-of-the-art methods, comparisons against whom are also presented. [ABSTRACT FROM PUBLISHER]

Many applications of autonomy are significantly complicated by the need for wireless networking, with challenges including scalability and robustness. Radio accomplishes this in a complex environment, but suffers from rapid signal strength variation and attenuation typically much worse than free space loss. In this paper, we propose and test algorithms to autonomously discover the connectivity area for a base station in an unknown environment using an average of received signal strength (RSS) values and a RSS threshold to delineate the goodness of the channel. We combine region decomposition and RSS sampling to cast the problem as an efficient graph search. The nominal RSS in a sampling region is obtained by averaging local RSS samples to reduce the small-scale fading variation. The RSS gradient is exploited during exploration to develop an efficient approach for discovery of the base station connectivity boundary in an unknown environment. Indoor and outdoor experiments demonstrate the proposed techniques. The results can be used for sensing and collaborative autonomy, building base station coverage maps in unknown environments, and facilitating multi-hop relaying to a base station. [ABSTRACT FROM AUTHOR]

Numerous approaches have been proposed to generate well-balanced gaits in biped robots that show excellent performance in simulated environments. However, in general, the dynamic balance of the robots decreases dramatically when these methods are tested in physical platforms. Since humanoid robots are intended to collaborate with humans and operate in everyday environments, it is of paramount importance to test such approaches both in physical platforms and under severe conditions. In this work, the special characteristics of the Nao humanoid platform are analyzed and a control system that allows robust walking and disturbance rejection is proposed. This approach combines the zero moment point (ZMP) stability criterion with angular momentum suppression and step timing control. The proposed method is especially suitable for platforms with limited computational resources and sensory and sensory-motor capabilities. [ABSTRACT FROM AUTHOR]

We describe an integrated strategy for planning, perception, state estimation and action in complex mobile manipulation domains based on planning in the belief space of probability distributions over states using hierarchical goal regression (pre-image back-chaining). We develop a vocabulary of logical expressions that describe sets of belief states, which are goals and subgoals in the planning process. We show that a relatively small set of symbolic operators can give rise to task-oriented perception in support of the manipulation goals. An implementation of this method is demonstrated in simulation and on a real PR2 robot, showing robust, flexible solution of mobile manipulation problems with multiple objects and substantial uncertainty. [ABSTRACT FROM AUTHOR]

We present the design of a novel compliant quadruped robot, called Cheetah-cub, and a series of locomotion experiments with fast trotting gaits. The robot’s leg configuration is based on a spring-loaded, pantograph mechanism with multiple segments. A dedicated open-loop locomotion controller was derived and implemented. Experiments were run in simulation and in hardware on flat terrain and with a step down, demonstrating the robot’s self-stabilizing properties. The robot reached a running trot with short flight phases with a maximum Froude number of FR = 1.30, or 6.9 body lengths per second. Morphological parameters such as the leg design also played a role. By adding distal in-series elasticity, self-stability and maximum robot speed improved. Our robot has several advantages, especially when compared with larger and stiffer quadruped robot designs. (1) It is, to the best of the authors’ knowledge, the fastest of all quadruped robots below 30kg (in terms of Froude number and body lengths per second). (2) It shows self-stabilizing behavior over a large range of speeds with open-loop control. (3) It is lightweight, compact, and electrically powered. (4) It is cheap, easy to reproduce, robust, and safe to handle. This makes it an excellent tool for research of multi-segment legs in quadruped robots. [ABSTRACT FROM AUTHOR]

The central challenge in robotic mapping is obtaining reliable data associations (or “loop closures”): state-of-the-art inference algorithms can fail catastrophically if even one erroneous loop closure is incorporated into the map. Consequently, much work has been done to push error rates closer to zero. However, a long-lived or multi-robot system will still encounter errors, leading to system failure. We propose a fundamentally different approach: allow richer error models that allow the probability of a failure to be explicitly modeled. In other words, rather than characterizing loop closures as being “right” or “wrong”, we propose characterizing the error of those loop closures in a more expressive manner that can account for their non-Gaussian behavior. Our approach leads to an fully integrated Bayesian framework for dealing with error-prone data. Unlike earlier multiple-hypothesis approaches, our approach avoids exponential memory complexity and is fast enough for real-time performance. We show that the proposed method not only allows loop closing errors to be automatically identified, but also that in extreme cases, the “front-end” loop-validation systems can be unnecessary. We demonstrate our system both on standard benchmarks and on the real-world data sets that motivated this work. [ABSTRACT FROM AUTHOR]

The development of legged robots for complex environments requires controllers that guarantee both high tracking performance and compliance with the environment. More specifically the control of the contact interaction with the environment is of crucial importance to ensure stable, robust and safe motions. In this contribution we develop an inverse-dynamics controller for floating-base robots under contact constraints that can minimize any combination of linear and quadratic costs in the contact constraints and the commands. Our main result is the exact analytical derivation of the controller. Such a result is particularly relevant for legged robots as it allows us to use torque redundancy to directly optimize contact interactions. For example, given a desired locomotion behavior, we can guarantee the minimization of contact forces to reduce slipping on difficult terrains while ensuring high tracking performance of the desired motion. The main advantages of the controller are its simplicity, computational efficiency and robustness to model inaccuracies. We present detailed experimental results on simulated humanoid and quadruped robots as well as a real quadruped robot. The experiments demonstrate that the controller can greatly improve the robustness of locomotion of the robots. [ABSTRACT FROM AUTHOR]

We develop robust methods that allow the specification, control, and transition of a multi-legged robot’s stepping pattern, its ‘gait’, during active locomotion over natural terrain. Resulting gaits emerge through the introduction of controllers that impose appropriately placed repellors within the space of gaits, the torus of relative leg phases, thereby mitigating against dangerous patterns of leg timing. Moreover, these repellors are organized with respect to a natural cellular decomposition of gait space and result in limit cycles with associated basins that are well characterized by these cells, thus conferring a symbolic character upon the overall behavioral repertoire. These ideas are particularly applicable to four- and six-legged robots, for which a large variety of interesting and useful (and, in many cases, familiar) gaits exist, and whose tradeoffs between speed and reliability motivate the desire for transitioning between them during active locomotion. We provide an empirical instance of this gait regulation scheme by application to a climbing hexapod, whose ‘physical layer’ sensor-feedback control requires adequate grasp of a climbing surface but whose closed-loop control perturbs the robot from its desired gait. We document how the regulation scheme secures the desired gait and permits operator selection of different gaits as required during active climbing on challenging surfaces. [ABSTRACT FROM PUBLISHER]

Variable stiffness actuators (VSAs) are currently explored as a new actuation approach to increase safety in physical human–robot interaction (pHRI) and improve dynamic performance of robots. For control purposes, accurate knowledge is needed of the varying stiffness at the robot joints, which is not directly measurable, nonlinearly depending on transmission deformation, and uncertain to be modeled. We address the online estimation of transmission stiffness in robots driven by VSAs in antagonistic or serial configuration, without the need for joint torque sensing. The two-stage approach combines (i) a residual-based estimator of the torque at the flexible transmission, and (ii) a recursive least squares stiffness estimator based on a parametric model. Further design refinements guarantee a robust behavior in the lack of velocity measures and in poor excitation conditions. The proposed stiffness estimation can be easily extended to multi-degree-of-freedom (multi-DOF) robots in a decentralized way, using only local motor and link position measurements. The method is tested through extensive simulations on the VSA-II device of the University of Pisa and on the Actuator with Adjustable Stiffness (AwAS) of IIT. Experiments on the AwAS platform validate the approach. [ABSTRACT FROM PUBLISHER]

Physical human–robot interaction implies the intersection of human and robot workspaces and intrinsically favors collision. The robustness of the most exposed parts, such as the hands, is crucial for effective and complete task execution of a robot. Considering the scales, we think that the robustness can only be achieved by the use of energy storage mechanisms, e.g. in elastic elements. The use of variable stiffness drives provides a low-pass filtering of impacts and allows stiffness adjustments depending on the task. However, using these drive principles does not guarantee the safety of the human due to the dramatically increased dynamics of such system. The design methodology of an antagonistically tendon-driven hand is explained. The resulting hand, very close to its human archetype in terms of size, weight, and, in particular, grasping performance, robustness, and dynamics, is presented. The hyper-actuated hand is a research platform that will also be used to investigate the importance of mechanical couplings and, in future projects, be the basis of a simplified hand that would still perform daily manipulation tasks. [ABSTRACT FROM PUBLISHER]

We present a 3D model-based visual tracking approach using edge and keypoint features in a particle filtering framework. Recently, particle-filtering-based approaches have been proposed to integrate multiple pose hypotheses and have shown good performance, but most of the work has made an assumption that an initial pose is given. To ameliorate this limitation, we employ keypoint features for initialization of the filter. Given 2D–3D keypoint correspondences, we randomly choose a set of minimum correspondences to calculate a set of possible pose hypotheses. Based on the inlier ratio of correspondences, the set of poses are drawn to initialize particles. After the initialization, edge points are employed to estimate inter-frame motions. While we follow a standard edge-based tracking, we perform a refinement process to improve the edge correspondences between sampled model edge points and image edge points. For better tracking performance, we employ a first-order autoregressive state dynamics, which propagates particles more effectively than Gaussian random walk models. The proposed system re-initializes particles by itself when the tracked object goes out of the field of view or is occluded. The robustness and accuracy of our approach is demonstrated using comparative experiments on synthetic and real image sequences. [ABSTRACT FROM PUBLISHER]

In this paper we present a visual navigation algorithm for the six-legged walking robot DLR Crawler in rough terrain. The algorithm is based on stereo images from which depth images are computed using the semi-global matching (SGM) method. Further, a visual odometry is calculated along with an error measure. Pose estimates are obtained by fusing inertial data with relative leg odometry and visual odometry measurements using an indirect information filter. The visual odometry error measure is used in the filtering process to put lower weights on erroneous visual odometry data, hence, improving the robustness of pose estimation. From the estimated poses and the depth images, a dense digital terrain map is created by applying the locus method. The traversability of the terrain is estimated by a plane fitting approach and paths are planned using a D* Lite planner taking the traversability of the terrain and the current motion capabilities of the robot into account. Motion commands and the traversability measures of the upcoming terrain are sent to the walking layer of the robot so that it can choose an appropriate gait for the terrain. Experimental results show the accuracy of the navigation algorithm and its robustness against visual disturbances. [ABSTRACT FROM PUBLISHER]

We present MOPED, a framework for Multiple Object Pose Estimation and Detection that seamlessly integrates single-image and multi-image object recognition and pose estimation in one optimized, robust, and scalable framework. We address two main challenges in computer vision for robotics: robust performance in complex scenes, and low latency for real-time operation. We achieve robust performance with Iterative Clustering Estimation (ICE), a novel algorithm that iteratively combines feature clustering with robust pose estimation. Feature clustering quickly partitions the scene and produces object hypotheses. The hypotheses are used to further refine the feature clusters, and the two steps iterate until convergence. ICE is easy to parallelize, and easily integrates single- and multi-camera object recognition and pose estimation. We also introduce a novel object hypothesis scoring function based on M-estimator theory, and a novel pose clustering algorithm that robustly handles recognition outliers. We achieve scalability and low latency with an improved feature matching algorithm for large databases, a GPU/CPU hybrid architecture that exploits parallelism at all levels, and an optimized resource scheduler. We provide extensive experimental results demonstrating state-of-the-art performance in terms of recognition, scalability, and latency in real-world robotic applications. [ABSTRACT FROM AUTHOR]

The planar bipedal testbed MABEL contains springs in its drivetrain for the purpose of enhancing both energy efficiency and agility of dynamic locomotion. While the potential energetic benefits of springs are well documented in the literature, feedback control designs that effectively realize this potential are lacking. In this paper, we extend and apply the methods of virtual constraints and hybrid zero dynamics, originally developed for rigid robots with a single degree of underactuation, to MABEL, a bipedal walker with a novel compliant transmission and multiple degrees of underactuation. A time-invariant feedback controller is designed such that the closed-loop system respects the natural compliance of the open-loop system and realizes exponentially stable walking gaits. Five experiments are presented that highlight different aspects of MABEL and the feedback design method, ranging from basic elements such as stable walking and robustness under perturbations, to energy efficiency and a walking speed of 1.5 m s−1 (3.4 mph). The experiments also compare two feedback implementations of the virtual constraints, one based on PD control of Westervelt et al., and a second that implements a full hybrid zero dynamics controller. On MABEL, the full hybrid zero dynamics controller yields a much more faithful realization of the desired virtual constraints and was instrumental in achieving more rapid walking. [ABSTRACT FROM AUTHOR]

Robots usually carry out object segmentation and modeling passively. Sensors such as cameras are actuated by a robot without disturbing objects in the scene. In this paper, we present an intelligent robotic system that physically moves objects in an active manner to perform segmentation and modeling using vision. By visually detecting bilateral symmetry, our robot is able to segment and model objects through controlled physical interactions. Extensive experiments show that our robot is able to accurately segment new objects autonomously. We also show that our robot is able to leverage segmentation results to autonomously learn visual models of new objects by physically grasping and rotating them. Object recognition experiments confirm that the robot-learned models allow robust recognition. Videos of the robotic experiments are also made available. [ABSTRACT FROM AUTHOR]

We describe a new formulation of appearance-only SLAM suitable for very large scale place recognition. The system navigates in the space of appearance, assigning each new observation to either a new or a previously visited location, without reference to metric position. The system is demonstrated performing reliable online appearance mapping and loop-closure detection over a 1000 km trajectory, with mean filter update times of 14 ms. The scalability of the system is achieved by defining a sparse approximation to the FAB-MAP model suitable for implementation using an inverted index. Our formulation of the problem is fully probabilistic and naturally incorporates robustness against perceptual aliasing. We also demonstrate that the approach substantially outperforms the standard term-frequency inverse-document-frequency (tf-idf) ranking measure. The 1000 km data set comprising almost a terabyte of omni-directional and stereo imagery is available for use, and we hope that it will serve as a benchmark for future systems. [ABSTRACT FROM AUTHOR]

We consider the problem of multi-robot task-allocation when robots have to deal with uncertain utility estimates. Typically an allocation is performed to maximize expected utility; we consider a means for measuring the robustness of a given optimal allocation when robots have some measure of the uncertainty (e.g. a probability distribution, or moments of such distributions). We introduce the interval Hungarian algorithm, a new algorithm that extends the classic Kuhn—Munkres Hungarian algorithm to compute the maximum interval of deviation, for each entry in the assignment matrix, which will retain the same optimal assignment. The algorithm has a worst-case time complexity of O(n4); we also introduce a parallel variant with O(n3) running time, which is able to exploit the concurrent computing capabilities of distributed multi-robot systems. This provides an efficient measurement of the tolerance of the allocation to the uncertainties and dynamics, for both a specific interval and a set of interrelated intervals. We conduct experiments both in simulation and with physical robots to validate the approach and to gain insight into the effect of location uncertainty on allocations for multi-robot multi-target navigation tasks. [ABSTRACT FROM AUTHOR]

One of the hallmarks of the performance, versatility, and robustness of biological motor control is the ability to adapt the impedance of the overall biomechanical system to different task requirements and stochastic disturbances. A transfer of this principle to robotics is desirable, for instance to enable robots to work robustly and safely in everyday human environments. It is, however, not trivial to derive variable impedance controllers for practical high degree-of-freedom (DOF) robotic tasks. In this contribution, we accomplish such variable impedance control with the reinforcement learning (RL) algorithm PI2 (PolicyImprovement withPath Integrals). PI2is a model-free, sampling-based learning method derived from first principles of stochastic optimal control. The PI2algorithm requires no tuning of algorithmic parameters besides the exploration noise. The designer can thus fully focus on the cost function design to specify the task. From the viewpoint of robotics, a particular useful property of PI2is that it can scale to problems of many DOFs, so that reinforcement learning on real robotic systems becomes feasible. We sketch the PI2algorithm and its theoretical properties, and how it is applied to gain scheduling for variable impedance control. We evaluate our approach by presenting results on several simulated and real robots. We consider tasks involving accurate tracking through via points, and manipulation tasks requiring physical contact with the environment. In these tasks, the optimal strategy requires both tuning of a reference trajectory and the impedance of the end-effector. The results show that we can use path integral based reinforcement learning not only for planning but also to derive variable gain feedback controllers in realistic scenarios. Thus, the power of variable impedance control is made available to a wide variety of robotic systems and practical applications. [ABSTRACT FROM AUTHOR]

In this paper we present a ‘fingerprint method’ for modeling and subsequently characterizing stochastically controlled actuator arrays. The actuator arrays are built from small actuator cells with structural elasticity. These cells are controlled using a bistable stochastic process wherein all cells are given a common input probability (control) value which they use to determine whether to actuate or relax. Arranging the cells in different networks gives different actuator array properties, which must be found before the actuator arrays can be applied to manipulators. The fingerprint method is used to describe and automatically generate every possible stochastic actuator array topology for a given number of cells, and to calculate actuator array properties such as: travel, required actuator strength/displacement, force range, force variance, and robustness for any array topology. The properties of several illustrative examples are shown and a discussion covers the importance of the properties, and trends between actuator array layouts and their properties. Finally, results from a validation experiment using a stochastically controlled solenoid array are presented. [ABSTRACT FROM AUTHOR]

With the increasing complexity of robotic systems, system robustness and efficiency are harder to achieve, since they are determined by the interplay of all of a system’s components. In order to improve the robustness of such systems, it is essential to identify the system components that are crucial for each task and the extent to which they are affected by other components and the environment. Such knowledge will help developers to improve their systems, and can also be directly utilized by the systems themselves, for example, to detect failures and thereby correctly adjust the system’s behavior.In this article a method of system interdependence analysis is presented. The basic idea is to learn and quantitatively evaluate the coherence between performance indicators of different system components, as well as the influence of environmental parameters on the system. To validate the proposed approach, system interdependence analysis is applied to the navigation system of an autonomous mobile robot. Its navigational methods are presented and suitable indicators are derived. The results of using the method, based on experimental data from an extended field experiment, are given. [ABSTRACT FROM AUTHOR]

Slip between the ground and wheel often cannot be avoided in most applications of mobile robots. However, a majority of controllers developed so far make a no-slip assumption with non-holonomic constraints. To achieve desired performance in the presence of slip, controllers that are robust to slip are required. In this paper, we discuss robust trajectory-tracking control for a differentially driven two-wheeled mobile robot. The structure of a differential flatness controller, which has shown distinctive advantages providing an integrated framework for planning and control, is extended to account for slip disturbances. It is shown that the differential flatness framework can be extended to develop a robust controller based on a dynamic as well as a kinematic model with slip. Simulation results for both kinematic and dynamic controllers are presented to demonstrate the effectiveness of the robust controllers. Experiments with the kinematic controller which is suited to typical laboratory and field mobile robots were conducted to validate the proposed robust controller. The simulation and experimental results show that the proposed robust controllers are effective in the presence of slip. [ABSTRACT FROM AUTHOR]

This paper presents the design and optimization of a wall-climbing robot along with the incorporation of autonomous adhesion recovery and a motion planning implementation. The result is Waalbot II, an untethered 85 g robot able to climb on smooth vertical surfaces with up to a 100 g payload (117% body mass) or, when unburdened, on planar surfaces of any orientation at speeds up to 5 cm/s. Bio-inspired climbing mechanisms, such as Waalbot II’s gecko-like fibrillar adhesives, passive peeling, and force sensing, improve the overall climbing capabilities compared with initial versions, resulting in the ability to climb on non-smooth surfaces as well as on inverted smooth surfaces. Robot length scale optimization reveals and quantifies trends in the theoretical factor of safety and payload carrying capabilities. Autonomous adhesion recovery behavior provides additional climbing robustness without additional mechanical complexity to mitigate degradation and contamination. An implementation of a motion planner, designed to take into account Waalbot II’s kinematic constraints, results in the ability to navigate to a goal in complex three-dimensional environments while properly planning plane-to-plane transitions and avoiding obstacles. Experiments verified the improved climbing capabilities of Waalbot II as well as its novel semi-autonomous adhesion recovery behavior and motion planning. [ABSTRACT FROM AUTHOR]

Rough terrain autonomous navigation continues to pose a challenge to the robotics community. Robust navigation by a mobile robot depends not only on the individual performance of perception and planning systems, but on how well these systems are coupled. When traversing complex unstructured terrain, this coupling (in the form of a cost function) has a large impact on robot behavior and performance, necessitating a robust design. This paper explores the application of Learning from Demonstration to this task for the Crusher autonomous navigation platform. Using expert examples of desired navigation behavior, mappings from both online and offline perceptual data to planning costs are learned. Challenges in adapting existing techniques to complex online planning systems and imperfect demonstration are addressed, along with additional practical considerations. The benefits to autonomous performance of this approach are examined, as well as the decrease in necessary designer effort. Experimental results are presented from autonomous traverses through complex natural environments. [ABSTRACT FROM AUTHOR]

In this paper, the roll and pitch dynamics of a biologically inspired quadruped water runner robot are analyzed, and a stable robot design is proposed and tested. The robot’s foot—water interaction force is derived using drag equations. Roll direction instability is attributed to a small roll moment of inertia and large instantaneous roll moments generated by the foot—water interaction forces. Roll dynamics are modeled by approximating the water as a linear spring. Using this model, estimates on the roll moment of inertia that can endure moments generated by water interactions are derived. Instability in the pitch direction is caused by the thrust force the four feet exert on the water. To correct this, a circular tail which can negate the pitch moment around the center of mass is proposed. Both passive and active tail designs which can cope with disturbances are introduced. Based on these analyses, a stable water runner is designed, and built. Experimental high-speed video footage demonstrates the stable roll and pitch motion of the robot. Simulations are used to estimate robustness against disturbances, waves, and leg running frequency variations. It is found that roll motion is more sensitive to disturbances when compared with the pitch direction. [ABSTRACT FROM AUTHOR]

Humanoid robots are expected to work in the human environment due to their similarity to the human shape, and they are required to achieve localization and navigation autonomously. In this paper, indoor navigation is realized based on a view-based approach using a camera mounted on the head. In the case of a humanoid robot, image blur and swing due to walking is a crucial issue for image matching during localization. The quantitative effect of walking on the image is firstly investigated by utilizing a motion capture system. Then a method to generate a stable view sequence is proposed based on the detection of optical flows. The navigation function with the proposed method was implemented on a humanoid robot HRP-2, and the effectiveness was confirmed by indoor walking experiments. [ABSTRACT FROM AUTHOR]

The development of optimized motions of humanoid robots that guarantee fast and also stable walking is an important task, especially in the context of autonomous soccer-playing robots in RoboCup. We present a walking motion optimization approach for the humanoid robot prototype HR18 which is equipped with a low-dimensional parameterized walking trajectory generator, joint motor controller and an internal stabilization. The robot is included as hardware-in-the-loop to define a low-dimensional black-box optimization problem. In contrast to previously performed walking optimization approaches, we apply a sequential surrogate optimization approach using stochastic approximation of the underlying objective function and sequential quadratic programming to search for a fast and stable walking motion. This is done under the conditions that only a small number of physical walking experiments should have to be carried out during the online optimization process. For the identified walking motion for the considered 55 cm tall humanoid robot, we measured a forward walking speed of more than 30 cm s-1. With a modified version of the robot, even more than 40 cms-1 could be achieved in permanent operation. [ABSTRACT FROM AUTHOR]

In this paper we introduce a robust descriptor for matching vertical lines among two or more images from an omnidirectional camera. Furthermore, in order to make such a descriptor usable in the framework of indoor mobile robotics, this paper introduces a new simple strategy to extrinsically self-calibrate the omnidirectional sensor with the odometry reference system. In the first part of this paper we describe how to build the feature descriptor. We show that the descriptor is very distinctive and is invariant to rotation and slight changes in illumination. The robustness of the descriptor is validated through real experiments on a wheeled robot. The second part of the paper is devoted to the extrinsic self-calibration of the camera with the odometry reference system. We show that by implementing an extended Kalman filter that fuses the information from the visual features with the odometry, it is possible to extrinsically and automatically calibrate the camera while the robot is moving. In particular, it is theoretically shown that only one feature suffices to perform the calibration. Experimental results validate the theoretical contributions. [ABSTRACT FROM AUTHOR]

Simultaneous Localization and Mapping (SLAM) is one of the classical problems in mobile robotics. The task is to build a map of the environment using on-board sensors while at the same time localizing the robot relative to this map. Rao-Blackwellized particle filters have emerged as a powerful technique for solving the SLAM problem in a wide variety of environments. It is a well-known fact for samplingbased approaches that the choice of the proposal distribution greatly influences robustness and efficiency achievable by the algorithm. In this paper, we present an improved proposal distribution for gridbased SLAM with Rao-Blackwellized particle filters, which utilizes whole sequences of sensor measurements rather than only the most recent one.We have implemented our system on a real robot and evaluated its performance on standard datasets as well as in hard outdoor settings with few and ambiguous features. Our approach improves the localization accuracy and the map quality, substantially reducing the risk of mapping failures. [ABSTRACT FROM AUTHOR]

The classical occupancy grid formulation requires the use of a priori known measurement likelihoods whose values are typically either assumed or learned from training data. Furthermore, in previous approaches, the likelihoods used to propagate the occupancy map variables are, in fact, independent of the state of interest and are derived from the spatial uncertainty of the detected point. This allows the use of a discrete Bayes filter as a solution to the problem, as discrete occupancy measurement likelihoods are used. In this paper, we first shown that once the measurement space is redefined, theoretically accurate and state-dependant measurement likelihoods can be obtained and used in the propagation of the occupancy random variable. The required measurement likelihoods for occupancy filtering are, in fact, those commonly encountered in both the landmark detection and data association hypotheses decisions. However, the required likelihoods are generally a priori unknown as they are a highly non-linear function of the landmark's signal-to-noise ratio and the surrounding environment. The probabilistic occupancy mapping problem is therefore reformulated as a continuous joint estimation problem where the measurement likelihoods are treated as continuous random states which must be jointly estimated with the map. In particular, this work explicitly considers the sensors detection and false-alarm probabilities in the occupancy mapping formulation. A particle solution is proposed which recursively estimates both the posterior on the map and the measurement likelihoods. The ideas presented in this paper are demonstrated in the field robotics domain using a millimeter wave radar sensor and comparisons with previous approaches, using constant discrete measurement likelihoods, are shown. A manually constructed ground-truth map and satellite imagery are also provided for map validation. [ABSTRACT FROM AUTHOR]

Dexterous manipulation with a highly redundant movement system is one of the hallmarks of human motor skills. From numerous behavioral studies, there is strong evidence that humans employ compliant task space control, i.e. they focus control only on task variables while keeping redundant degrees-of-freedom as compliant as possible. This strategy is robust towards unknown disturbances and simultaneously safe for the operator and the environment. The theory of operational space control in robotics aims to achieve similar performance properties. However despite various compelling theoretical lines of research, advanced operational space control is hardly found in actual robotics implementations, in particular new kinds of robots like humanoids and service robots, which would strongly profit front com- pliant dexterous manipulation. To analyze the pros and cons of different approaches to operational space control, this paper focuses on a theoretical and empirical evaluation of different methods that have been suggested in the literature, but also some new variants of operational space controllers. We address formulations at the velocity; acceleration, and force levels. First, we formulate all controllers in a common notational framework, including quaternion-based orientation control, and discuss some of their theoretical properties. Second, we present experimental comparisons of these approaches on a seven-degree-of-freedom anthropomorphic robot arm with several benchmark tasks. As an aside, we also introduce a novel parameter estimation algorithm for rigid body dynamics, which ensures physical consistency, as this issue was crucial for our successful robot implementations. Our extensive empirical results demonstrate that one of the simplified acceleration-based approaches can be advantageous in terms of task performance, ease of parameter tuning, and general robustness and compliance in the face of inevitable modeling errors. [ABSTRACT FROM AUTHOR]

In this paper a new robust fault detection technique for robotic manipulators is developed. The new approach, called robust nonlinear analytic redundancy (RNLAR) technique, detects both sensor and actuator faults in a robotic manipulator The proposed RNLAR technique can compensate for the effects of model—plan -mismatch (MPM) and process disturbance. A nonlinear primary residual vectors (PRV) design method to detect faults is proposed where the PRVs are highly sensitive to faults and less sensitive to MPM and process disturbance. Experimental results on a PUMA 560 are presented to justify the effectiveness of the RNLAR scheme. [ABSTRACT FROM AUTHOR]

Stability is essential for teleoperation and a prerequisite for performance. This paper analyzes the the stability/performance trade-off of a teleoperator where the slave device has a built-in passive intrinsic stiffness. Stability is quantified as time delay robustness and performance is expressed using teleoperator damping and teleoperator stiffness, the boundaries of the Colgate Z-width. Two classic control schemes, position error and Lawrence 4-channel, are used along with a novel 5-channel scheme where the slave stiffness deflection is measured, and compensated for to improve the performance. The teleoperator system was analyzed theoretically using a linear model and the findings were experimentally validated on a one degree of freedom teleoperation setup. It was found that: • A lower slave stiffness improves stability for all three teleoperator architectures. • The stability boundary of the three controllers is similar. • The performance of the controllers increases from: (poor) position error 4-Channel to (excellent) 5-channel. • A classical linear analysis method can accurately predict the stability characteristics of the teleoperator system. Therefore it can be concluded that a compliant slave device offers a stability advantage for a range of teleoperation situations and that the loss of performance can be partly compensated. [ABSTRACT FROM AUTHOR]

This paper is concerned with computation and graphical characterization of robust equilibrium postures suited to quasistatic multi-legged locomotion. Quasistatic locomotion consists of postures ill which the mechanism supports itself against gravity while moving its free limbs to new positions. A posture is robust if the contacts can passively support the mechanism against gravity as well as disturbance forces generated by its moving limbs. This paper is concerned with planar mechanisms supported by frictional contacts in two-dimensional gravitational environments. The kinematic structure of the mechanism is lumped into a rigid body B having the same contacts with the environment and a variable center of mass. Inertial forces generated by moving parts of the mechanism are lumped into a neighborhood of wrenches centered at the nominal gravitational wrench. The robust equilibrium postures associated with a given set of contacts become the center-of-mass locations of B that maintain equilibrium with respect to all wrenches in the given neighborhood. The paper formulates the computation of the robust center-of-mass locations as a linear programming problem. It provides graphical characterization of the robust center-of-mass locations, and gives a geometric algorithm for computing these center-of-mass locations. The paper reports experiments validating the equilibrium criterion on a two-legged prototype. Finally, it describes initial progress toward computation of robust equilibrium postures in three dimensions. [ABSTRACT FROM AUTHOR]

This paper presents the main simulation and experimental results from studies of the robustness of a proposed new control strategy for the under-actuated robot RABBIT The disturbances studied include modifications of the characteristics of the foot/ground interaction and the application of external forces to the trunk of the robot. These two kinds of disturbance correspond to those most frequently encountered by biped robots during walking in an unknown environment. [ABSTRACT FROM AUTHOR]