ICRA 2015 论文列表

This paper considers the problem of approximating a kernel matrix in an autoregressive Gaussian process regression (AR-GP) in the presence of measurement noises or natural errors for modeling complex motions of pedestrians in a crowded environment. While a number of methods have been proposed to robustly predict future motions of humans, it still remains as a difficult problem in the presence of measurement noises. This paper addresses this issue by proposing a structured low-rank matrix approximation method using nuclear-norm regularized l1-norm minimization in AR-GP for robust motion prediction of dynamic obstacles. The proposed method approximates a kernel matrix by finding an orthogonal basis using low-rank symmetric positive semi-definite matrix approximation assuming that a kernel matrix can be well represented by a small number of dominating basis vectors. The proposed method is suitable for predicting the motion of a pedestrian, such that it can be used for safe autonomous robot navigation in a crowded environment. The proposed method is applied to well-known regression and motion prediction problems to demonstrate its robustness and excellent performance compared to existing approaches.

Within this paper, a new fast algorithm that provides efficient solutions to the problem of inspection path planning for complex 3D structures is presented. The algorithm assumes a triangular mesh representation of the structure and employs an alternating two-step optimization paradigm to find good viewpoints that together provide full coverage and a connecting path that has low cost. In every iteration, the viewpoints are chosen such that the connection cost is reduced and, subsequently, the tour is optimized. Vehicle and sensor limitations are respected within both steps. Sample implementations are provided for rotorcraft and fixed-wing unmanned aerial systems. The resulting algorithm characteristics are evaluated using simulation studies as well as multiple real-world experimental test-cases with both vehicle types.

An autonomous robot team can be employed for continuous and strategic coverage of arbitrary environments for different missions. In this work, we propose an anytime approach for creating multi-robot patrolling policies. Our approach involves a novel extension of Monte Carlo Tree Search (MCTS) to allow robots to have life-long, cyclic policies so as to provide continual coverage of an environment. Our proposed method can generate near-optimal policies for a team of robots for small environments in real-time (and in larger environments in under a minute). By incorporating additional planning heuristics we are able to plan coordinated patrolling paths for teams of several robots in large environments quickly on commodity hardware.

Many species in nature store information in the environment to facilitate the performance of tasks and enable cooperation. This principle is known as stigmergy. Stigmergy has been widely studied in robotic systems, but so far mostly in simulation or in laboratory proofs of concept. In this paper, we propose a stigmergic approach to goal-directed navigation that can be used for navigation of a full-scale robotic system in a real apartment. A team of small ePuck robots build a set of navigation maps directly onto an RFID floor, where each map is associated to one predefined goal. The information stored in the floor can then used by a mid-size robot or by a larger domestic robot to perform safe navigation toward the predefined goals. To navigate, robots only rely on the information read from the RFID tags: in particular, they do not need to use an internal map or to perform self-localization. This results in robust and repeatable navigation with minimal hardware and software requirements.

Dynamic system states estimation, such as object pose and contact states estimation, is essential for robots to perform manipulation tasks. In order to make accurate estimation, the state transition model needs to be physically correct. Complementarity formulations of the dynamics are widely used for describing rigid body physical behaviors in the simulation field, which makes it a good state transition model for dynamic system states estimation problem. However, the non-smoothness of complementarity models and the high dimensionality of the dynamic system make the estimation problem challenging. In this paper, we propose a particle filtering framework that solves the estimation problem by sampling the discrete contact states using contact graphs and collision detection algorithms, and by estimating the continuous states through a Kalman filter. This method exploits the piecewise continuous property of complementarity problems and reduces the dimension of the sampling space compared with sampling the high dimensional continuous states space. We demonstrate that this method makes stable and reliable estimation in physical experiments.

In grasping tasks carried out with humanoids, knowledge about the robot's reachable workspace is important. Without this knowledge, it might be necessary to repeatedly adapt the stance location and call an inverse kinematics solver before a valid robot configuration to reach a given grasping pose can be found. In this paper, we present an approach to select an optimal stance location in SE(2) for a humanoid robot's feet relative to a desired grasp pose. We use a precomputed representation of the robot's reachable workspace that stores quality information in addition to spatial data. By inverting this representation we obtain a so-called inverse reachability map (IRM) containing a collection of potential stance poses for the robot. The generated IRM can subsequently be used to select a statically stable, collision-free stance configuration to reach a given grasping target. We evaluated our approach with a Nao humanoid in simulation and in experiments with the real robot. As the experiments show, using our approach optimal stance poses can easily be obtained. Furthermore, the IRM leads to a substantially increased success rate of reaching grasping poses compared to other meaningful foot placements within the vicinity of the desired grasp.

Quadruped locomotion offers significant advantages over wheeled locomotion for small mobile robots operating in challenging terrain. Central pattern generators (CPGs), as found in the neural circuitry of many animals, may be used to generate joint trajectories for quadruped robots. However, basic CPG-based trajectories do not explicitly consider ground contact constraints, a particular concern during turning maneuvers when foot slip is most likely to occur. An alternative approach proposed here is to use task-based CPGs such that ground contact constraints are enforced and foot velocities are explicitly controlled, resulting in stable omnidirectional locomotion. Further, incorporating Virtual Model Control with the task-based CPG trajectories improves the stability of the quadruped in hardware experiments.

Humanoids locomote by making and breaking contacts with their environment. Thus, a crucial question for them is to anticipate whether a contact will hold or break under effort. For rigid surface contacts, existing methods usually consider several point-contact forces, which has some drawbacks due to the underlying redundancy. We derive a criterion, the Contact Wrench Cone (CWC), which is equivalent to any number of applied forces on the contact surface, and for which we provide a closed-form formula. It turns out that the CWC can be decomposed into three conditions: (i) Coulomb friction on the resultant force, (ii) CoP inside the support area, and (iii) upper and lower bounds on the yaw torque. While the first two are well-known, the third one is novel. It can, for instance, be used to prevent the undesired foot yaws observed in biped locomotion. We show that our formula yields simpler and faster computations than existing approaches for humanoid motions in single support, and assess its validity in the OpenHRP simulator.

Event-driven vision sensors have the potential to support a new generation of efficient and robust robots. This requires the development of a new computational framework that exploits not only the spatial information, like in the traditional frame-based approach, but also the temporal content of the sensory data. We propose a method for unsupervised learning of filters for the processing of the visual signal from event-driven sensors. This method exploits the temporal coincidence of events generated by each object in a spatial location of the visual field. The approach is based on a modification of Spike Timing Dependent Plasticity that takes into account the specific implementation on the robot and the characteristics of the used sensor. It gives rise to oriented spatial filters that are very similar to the receptive fields observed in the primary visual cortex and traditionally used in bio-inspired hierarchical structures for object recognition, as well as to novel curved spatial structures. Using mutual information measure we provide a quantitative evidence that such curved spatial filters provide more information than equivalent oriented Gabor filters and can be an important aspect for object recognition in robotic applications.

Experience Graphs have been shown to accelerate motion planning using parts of previous paths in an A* framework. Experience Graphs work by computing a new heuristic for weighted A* search on top of the domain's original heuristic and the edges in an Experience Graph. The new heuristic biases the search toward relevant prior experience and uses the original heuristic for guidance otherwise. In previous work, Experience Graphs were always built on top of domain heuristics which were computed by dynamic programming (a lower dimensional version of the original planning problem). When the original heuristic is computed this way the Experience Graph heuristic can be computed very efficiently. However, there are many commonly used heuristics in planning that are not computed in this fashion, such as euclidean distance. While the Experience Graph heuristic can be computed using these heuristics, it is not efficient, and in many cases the heuristic computation takes much of the planning time. In this work, we present a more efficient way to use these heuristics for motion planning problems by making use of popular nearest neighbor algorithms. Experimentally, we show an average 8 times reduction in heuristic computation time, resulting in overall planning time being reduced by 66%. with no change in the expanded states or resulting path.

In this paper, we present a new forward kinematic model for a novel class of long and thin continuum robots for operation in spatial workspace. Such robots are well suited for navigation through unstructured environments with superior reach using their flexible and thin profile, especially for inspection applications. This cable-like continuum robot design [1] has a concentric backbone arrangement but is spring-loaded and tendon-actuated, has improved compliance with a light and compact motor-encoder actuation mechanism. To account for the spring-loading, a compression factor is introduced on top of the established constant curvature continuum kinematics. The resulting continuum or shape variables are then estimated as a function of the measurable encoder variables. The effectiveness of the model is validated by performing experiments with the robot prototype.

Sparse Gaussian process (GP) models provide an efficient way to perform regression on large data sets. The key idea is to select a representative subset of the available training data, which induces the sparse GP model approximation. In the past, a variety of selection criteria for GP approximation have been proposed, but they either lack accuracy or suffer from high computational costs. In this paper, we introduce a novel and straightforward criterion for successive selection of training points used for GP model approximation. The proposed algorithm allows a fast and efficient selection of training points, while being competitive in learning performance. As evaluation, we employ our approach in learning inverse dynamics models for robot control using very large data sets (e.g. 500.000 samples). It is demonstrated in experiments that our approximated GP model is sufficiently fast for real-time prediction in robot control. Comparisons with other state-of-the-art approximation techniques show that our proposed approach is significantly faster, while being competitive to generalization accuracy.

This paper presents Sparse Depth Odometry (SDO) to incrementally estimate the 3D pose of a depth camera in indoor environments. SDO relies on 3D keypoints extracted on dense depth data and hence can be used to augment the RGB-D camera based visual odometry methods that fail in places where there is no proper illumination. In SDO, our main contribution is the design of the keypoint detection module, which plays a vital role as it condenses the input point cloud to a few keypoints. SDO differs from existing depth alone methods as it does not use the popular signed distance function and can run online, even without a GPU. A new keypoint detection module is proposed via keypoint selection, and is based on extensive theoretical and experimental evaluation. The proposed keypoint detection module comprises of two existing keypoint detectors, namely SURE [1] and NARF [2]. It offers reliable keypoints that describe the scene more comprehensively, compared to others. Finally, an extensive performance evaluation of SDO on benchmark datasets with the proposed keypoint detection module is presented and compared with the state-of-the-art.

This paper presents advancements in the design of a portable, soft robotic glove for individuals with functional grasp pathologies. The robotic glove leverages soft material actuator technology to safely distribute forces along the length of the finger and provide active flexion and passive extension. These actuators consist of molded elastomeric bladders with anisotropic fiber reinforcements that produce specific bending, twisting, and extending trajectories upon fluid pressurization. In particular, we present a method for customizing a soft actuator to a wearer's biomechanics and demonstrate in a motion capture system that the ranges of motion (ROM) of the two are nearly equivalent. The active ROM of the glove is further evaluated using the Kapandji test. Lastly, in a case study, we present preliminary results of a patient with very weak hand strength performing a timed Box-and-Block test with and without the soft robotic glove.

Mandibular mobility plays a significant role in human daily lives, enabling food intake, respiration, speaking, and other oral activities. However, as for the patients suffering from mandibular mobility disorders, their mandible functions are deteriorated, which severely affects their quality of life. In this paper, we present a new solution to recover mandibular mobility: a soft oral rehabilitation robot (SORR), which is actuated by a novel type of soft pneumatic actuator (SPA). After identifying the biometrics of the human mandible, we illustrate the application-oriented design of the robot and the SPA. A static model of the SPA was established to predict its behavior and eligibility for the application. In the experimental characterization, we measured the elongation and force output of the SPA, and the detailed results are presented. The comparison and analysis of the results provide physical insight into the mechanisms of the SPA. As the first application of soft robot inside human body, this work enlightens profound application potentials of the silicone-based SPAs.

The snakeboard is a well-studied example for mechanical systems analysis, largely because of its simultaneous richness in behavior and simplicity in design. However, few snakeboard models incorporate dissipative friction in the traveling direction and skidding as a violation of the rigid nonholonomic constraints. In this paper we investigate these effects on trajectory planning by evaluating a previously proposed friction model as well as a novel skidding model based on the addition of Rayleigh dissipation functions. We show how these additions change the usual behavior of gaits in the forward planning problem, and incorporate the changes into the solutions of the inverse planning problem by utilizing body coordinates along with a curvature parameterization for trajectories.

Encounter-type devices have been introduced to solve one of the main drawbacks of common haptic interfaces i.e. their limited transparency in free space and their restricted ability to render realistic transitions between free space and contact. Advantageously, encounter-type interfaces collide with fingers and display interaction forces only if a contact occurs in the virtual environment. In free space, their end-effectors are mechanically detached from the operator and closely track his fingers, allowing for a perfect transparency. This paper introduces a control scheme that handles the transition between tracking and force display modes in a smooth way. Experimental results using a custom encounter-type interface are also given. The obtained results show the efficiency of the proposed control law when compared to the conventional switch-based control usually found in the literature.

In this paper, we address the problem of stabilisation of robots subject to nonholonommic constraints and external disturbances using port-Hamiltonian theory and smooth time-invariant control laws. This should be contrasted with the commonly used switched or time-varying laws. We propose a control design that provides asymptotic stability of an manifold (also called relative equilibria)—due to the Brockett condition this is the only type of stabilisation possible using smooth time-invariant control laws. The equilibrium manifold can be shaped to certain extent to satisfy specific control objectives. The proposed control law also incorporates integral action, and thus the closed-loop system is robust to unknown constant disturbances. A key step in the proposed design is a change of coordinates not only in the momentum, but also in the position vector, which differs from coordinate transformations previously proposed in the literature for the control of nonholonomic systems. The theoretical properties of the control law are verified via numerical simulation based on a robotic ground vehicle model with differential traction wheels and non co-axial centre of mass and point of contact.

This paper describes the design, construction, and operation of a closed-loop spherical induction motor (SIM) ball wheel for a balancing mobile robot (ballbot). Following earlier work, this new design has a smaller rotor and higher torques due to the use of six stators in a skewed layout. Actuation and sensing kinematics as well as control methods are presented. In its current implementation, torques of up to 8 Nm are produced by the motor with rise and decay times of 100 ms. Results are presented supporting its potential as a prime mover for mobile robots.