IROS 2018 论文列表

Alignment of two point clouds is an essential problem in medical robotics and computer-assisted surgery. In this paper, we first formally formulate the generalized point cloud registration problem in a probabilistic manner. Specifically, not only positional but also the orientational information are incorporated into registration. Notably, the positional error is assumed to obey a multivariate Gaussian distribution to accommodate anisotropic cases. Expectation conditional maximization framework is utilized to solve the problem. In E-step, the correspondence probabilities between points in two generalized point clouds are computed. In M -step, the constrained optimization problem with respect to the transformation matrix is re-formulated as an unconstrained one. Extensive experiments are conducted to compare the proposed algorithm with the state-of-the-art registration methods. The experimental results demonstrate the algorithm's robustness to noise and outliers, fast convergence speed.

We present a novel fruit counting pipeline that combines deep segmentation, frame to frame tracking, and 3D localization to accurately count visible fruits across a sequence of images. Our pipeline works on image streams from a monocular camera, both in natural light, as well as with controlled illumination at night. We first train a Fully Convolutional Network (FCN) and segment video frame images into fruit and non-fruit pixels. We then track fruits across frames using the Hungarian Algorithm where the objective cost is determined from a Kalman Filter corrected Kanade-Lucas-Tomasi (KLT) Tracker. In order to correct the estimated count from tracking process, we combine tracking results with a Structure from Motion (SfM) algorithm to calculate relative 3D locations and size estimates to reject outliers and double counted fruit tracks. We evaluate our algorithm by comparing with ground-truth human-annotated visual counts. Our results demonstrate that our pipeline is able to accurately and reliably count fruits across image sequences, and the correction step can significantly improve the counting accuracy and robustness. Although discussed in the context of fruit counting, our work can extend to detection, tracking, and counting of a variety of other stationary features of interest such as leaf-spots, wilt, and blossom.

This paper presents a guidance vector field strategy to control a fixed-wing UAV (unmanned aerial vehicle)subject to uncertainty in order to converge to and circulate a closed curve in ℝ3. The control system is designed based on a reference model of the airplane with constrained input controls. The law is independent of the vector field's structure, however, some analysis considers a consolidated vector field approach. Asymptotic stability is proven with Lyapunov Theory and ultimate bounds are found when bounded uncertainties are taken into account. The control law is continuous except in the surroundings of the unavoidable field's singularities. A theorem ensures asymptotic convergence when a switch is made. Simulations with a 6 DOF, 12 states realistic aircraft model demonstrate the efficiency of the strategy and its advantages.

We study the ambiguous data association problem confronting simultaneous localization and mapping (SLAM), specifically for the autonomous exploration of environments lacking rich features. In such environments, a single false positive assignment might lead to catastrophic failure, which even robust back-ends may be unable to resolve. Inspired by multiple hypothesis tracking, we present a novel approach to effectively manage multiple hypotheses (MH) of data association inherited from traditional joint compatibility branch and bound (JCBB), which entails the generation, ordering and elimination of hypotheses. We analyze the performance of MHJCBB in two particular situations, one applying it to SLAM over a predefined trajectory and the other showing its applicability in exploring unknown environments. Statistical results demonstrate that MHJCBB's maintenance of diverse hypotheses under ambiguous conditions significantly improves map accuracy.

There is a high demand in advanced fault detection methods suitable for sensor networks monitoring complex dynamic systems such as industrial plants or large infrastructure units. This paper proposes a robust and efficient decentralized sensor fault detection method with in-place sensor self-recalibration capability that extracts and uses complex context information referred to the full monitored process. The method includes three main components, all decentralized and sharing the same statistical framework: 1) a consensus-based modeling step based on decentralized RANSAC; 2) a statistical analysis based on Bayesian networks and Hidden Markov Models in which each sensor identifies inconsistencies with the consensus model and determines if it is correctly calibrated, uncalibrated or faulty and; 3) a final step in which each uncalibrated sensor self-recalibrates using the consensus model. The proposed method is efficient in the use of computational and communicational resources, it is scalable and robust against outliers, transmission errors, sensor failures and network topology changes. It has been extensively validated in an experimental industrial setting.

Deep-sea robot operations demand a high level of safety, efficiency and reliability. As a consequence, measures within the development stage have to be implemented to extensively evaluate and benchmark system components ranging from data acquisition, perception and localization to control. We present an approach based on high-fidelity simulation that embeds spatial and environmental conditions from recorded real-world data. This simulation in the loop (SIL) methodology allows for mitigating the discrepancy between simulation and real-world conditions, e.g. regarding sensor noise. As a result, this work provides a platform to thoroughly investigate and benchmark behaviors of system components concurrently under real and simulated conditions. The conducted evaluation shows the benefit of the proposed work in tasks related to perception and self-localization under changing spatial and environmental conditions.

Estimating the camera pose requires point correspondences. However, in practice, correspondences are inevitably corrupted by outliers, which affects the pose estimation. We propose a general and accurate outlier removal strategy for robust camera pose estimation. The proposed strategy can detect outliers by leveraging the fact that only inliers comply with two effective consensuses, i.e., 3D ray bundle consensus and 2D vector field consensus. Our strategy has a nested structure. First, the outer module utilizes the 3D ray bundle consensus. We define the likelihood based on the probabilistic mixture model and maximize it by the expectation-maximization (EM) algorithm. The inlier probability of each correspondence and the camera pose are determined alternately. Second, the inner module exploits the 2D vector field consensus to refine the probabilities obtained by the outer module. The refinement based on the Bayesian rule facilitates the convergence of the outer module and improves the accuracy of the entire framework. Our strategy can be integrated into various existing camera pose estimation methods which are originally vulnerable to outliers. Experiments on both synthesized data and real images have shown that our approach outperforms state-of-the-art outlier rejection methods in terms of accuracy and robustness.

Object pose estimation is a crucial prerequisite for robots to perform autonomous manipulation in clutter. Real-world bin-picking settings such as warehouses present additional challenges, e.g., new objects are added constantly. Most of the existing object pose estimation methods assume that 3D models of the objects is available beforehand. We present a pipeline that requires minimal human intervention and circumvents the reliance on the availability of 3D models by a fast data acquisition method and a synthetic data generation procedure. This work builds on previous work on semantic segmentation of cluttered bin-picking scenes to isolate individual objects in clutter. An additional network is trained on synthetic scenes to estimate object poses from a cropped object-centered encoding extracted from the segmentation results. The proposed method is evaluated on a synthetic validation dataset and cluttered realworld scenes.

We propose a novel subsurface pipeline mapping method by fusing Ground Penetrating Radar (GPR) scans and camera images. To facilitate the simultaneous detection of multiple pipelines, we model the GPR sensing process and prove hyperbola response for general scanning with non-perpendicular angles. Furthermore, we fuse visual simultaneous localization and mapping outputs, encoder readings with GPR scans to classify hyperbolas into different pipeline groups. We extensively apply the J-Linkage method and maximum likelihood estimation to improve algorithm robustness and accuracy. As the result, we optimally estimate the radii and locations of all pipelines. We have implemented our method and tested it in physical experiments with representative pipeline configurations. The results show that our method successfully reconstructs all subsurface pipes. Moreover, the average localization error is 4.69cm.

This paper presents a versatile robotic system for sewing a 3D structured object. Leveraging on using a customized robotic sewing device and closed-loop visual servoing control, an all-in-one solution for sewing personalized stent graft is demonstrated. Stitch size planning and automated knot tying are proposed as two key functions of the system. By using effective stitch size planning, sub-millimetre sewing accuracy is achieved for stitch sizes ranging from 2mm to 5mm. In addition, a thread manipulator for thread management and tension control is also proposed to perform successive knot tying to secure each stitch. Detailed laboratory experiments have been performed to evaluate the proposed instruments and allied algorithms. The proposed framework can be generalised to a wide range of applications including 3D industrial sewing, as well as transferred to other clinical areas such as surgical suturing.

The robotic handling of compliant and deformable food raw materials, characterized by high biological variation, complex geometrical 3D shapes, and mechanical structures and texture, is currently in huge demand in the ocean space, agricultural, and food industries. Many tasks in these industries are performed manually by human operators who, due to the laborious and tedious nature of their tasks, exhibit high variability in execution, with variable outcomes. The introduction of robotic automation for most complex processing tasks has been challenging due to current robot learning policies. A more consistent learning policy involving skilled operators is desired. In this paper, we address the problem of robot learning when presented with inconsistent demonstrations. To this end, we propose a robust learning policy based on Learning from Demonstration (LfD) for robotic grasping of food compliant objects. The approach uses a merging of RGB-D images and tactile data in order to estimate the necessary pose of the gripper, gripper finger configuration and forces exerted on the object in order to achieve effective robot handling. During LfD training, the gripper pose, finger configurations and tactile values for the fingers, as well as RGB-D images are saved. We present an LfD learning policy that automatically removes inconsistent demonstrations, and estimates the teacher's intended policy. The performance of our approach is validated and demonstrated for fragile and compliant food objects with complex 3D shapes. The proposed approach has a vast range of potential applications in the aforementioned industry sectors.

This paper proposes a wearable robotic stick for walking assistance, called “Hand-Free-Stick” (HFS), for people with non-serious dysfunction in their gait. The basic idea of the proposed HFS is to enlarge ZMP (Zero moment point) area of a user under hands free conditions and to augment his/her body balance ability in walking. A boots type prototype of the HFS is developed with a lightweight robotic stick using a servomotor, in which the slider-link mechanism works to regulate the stick angle and length at the same time. The stick motion is controlled by a single-board computer based on the distribution of foot/feet pressures measured by the sensor system using eight load cells attached at the sole of boots. A set of walking tests with/without the prototype of HFS is carried out for four healthy subjects and demonstrates the effectiveness of the proposed HFS to expand the ZMP area leading to walking balance assistance.

The robustness of the positioning and posturing of robot hands relative to target object and support surface is an important issue for autonomous grasping. For example, to perform a grasping action such as picking up thin objects from a table top, the position and posture of the hand must be controlled to keep adequate relative posture and distance to the support surface besides those between the hand and the target object. Because slight errors in the posture and position are enough to cause grasping failure, the positioning and posturing of the hand must be precise enough, specially when the hand is close to the target object and support surface. To improve the robustness of robotic grasping, in this paper we present a method by grasping control based on the relative posture and position between hand and support surface besides those between hand and target object, using proximity sensors. Proximity sensors are newly installed on fingernails besides on the fingertips. As the fingernail sensor, an integration of Time-of-Flight (TOF) sensor and photo-reflector is designed to realize long range detection, as well as with precise and high-speed detection regardless of the reflectance of support surfaces when approaching the support surface. By the sensors, the hand can approach the object and support surface coarsely first, and then can be controlled fast and precisely to realize adequate grasping motion along the support surface but without contact with the support face. The method has been implemented to a manipulator system, and successful grasping experiments have demonstrated the effectiveness of the proposed method.

Automation of inspection and repair tasks on complex installations is gaining attention from industries with high-value assets such as aerospace, nuclear and marine. This paper reports on a five degrees of freedom robotic system capable of performing accurate and repeatable repair procedures through a narrow inspection port, which minimizes the cost and downtime associated with unscheduled maintenance. Careful study of the target working volume and repair process informed the design of a robotic probe capable of replicating the operation. Kinematic analysis of the robot's flexible, prismatic and rotary joints was used to define accurate machining paths in 3D space, and the results were verified using an optical motion capture system (accuracy of 0.25 mm). After comprehensive verifications of the constitutive elements, the robotic system was successfully demonstrated for repair of a high-pressure compressor aerofoil in a gas turbine. The results not only proves the ability of the system to address such difficult repair scenarios but also highlights a domain of opportunities in developing specialist robotics for repair of high-value assets, which is a subject to growing global demand.

Kinesthetic learning is a type of learning from demonstration in which the teacher manually moves the robot through the demonstrated trajectory. It shows great promise in the area of assistive robotics since it enables a caretaker who is not an expert in computer programming to communicate a novel task to an assistive robot. However, the trajectory the caretaker demonstrates to solve the task may be a high-cost trajectory for the robot. The demonstrated trajectory could be high-cost because the teacher does not know what trajectories are easy or hard for the robot to perform, which would be due to a limitation of the teacher's knowledge, or because the teacher has difficulty moving all the robotic joints precisely along the desired trajectories, which would be due to a limitation of the teacher's coordination. We propose the Parameterized Similar Path Search (PSPS) algorithm to extend kinesthetic learning so that a robot can improve the learned trajectory over a known cost function. This algorithm is based on active learning from the robot through collaboration between the robot's knowledge of the cost function and the caretaker's knowledge of the constraints of the assigned task.

Multiplayer environments are thought to increase and prolongate active participation in robot-aided rehabilitation. We expect that environments linking patients with their spouses will particularly foster active participation. Thus, we developed two multiplayer games to link the game experience of two players: an Air Hockey game and a Haptic Kitchen game. In the competitive Air Hockey game, differences in skill levels between players were balanced by individualizing haptic guidance or damping forces. In the Haptic Kitchen game, a healthy player could support the patient's movements using a virtual force field. The two players could control the haptic interaction since both the force field and the point of application were visualized. We tested the haptic performance balancing algorithm of the Air Hockey game and the spouse-controlled haptic support of the Kitchen game with patients post-stroke who trained both single- (i.e., alone) and multiplayer training (i.e., with spouse) in eight therapy sessions lasting 45 min each. Mean total rating in Intrinsic Motivation Inventory was 46.9 points (out of 63 points) for multiplayer modes, and 42.7 points for single player modes, respectively. The spouses applied the haptic support in the Haptic Kitchen game during 42 % of the total game duration. We are currently testing more patient-spouse couples to better understand the effects of using these haptic approaches on the behavior and recovery of patients. We foresee this approach can improve the motivation during training and positively influence the at-home behavior of patients, an important goal of rehabilitation training efforts.

This paper presents a future-focused approach for robot programming based on augmented trajectories. Using a mixed reality head-mounted display (Microsoft Hololens) and a 7-DOF robot arm, we designed an augmented reality (AR) robotic interface with four interactive functions to ease the robot programming task: 1) Trajectory specification. 2) Virtual previews of robot motion. 3) Visualization of robot parameters. 4) Online reprogramming during simulation and execution. We validate our AR-robot teaching interface by comparing it with a kinesthetic teaching interface in two different scenarios as part of a pilot study: creation of contact surface path and free space path. Furthermore, we present an industrial case study that illustrates our AR manufacturing paradigm by interacting with a 7-DOF robot arm to reduce wrinkles during the pleating step of the carbon-fiber-reinforcement-polymer vacuum bagging process in a simulated scenario.

Continuous Goal-Directed Actions (CGDA)is a robot imitation framework that encodes actions as the changes they produce on the environment. While it presents numerous advantages with respect to other robot imitation frameworks in terms of generalization and portability, final robot joint trajectories for the execution of actions are not necessarily encoded within the model. This is studied as an optimization problem, and the solution is computed through evolutionary algorithms in simulated environments. Evolutionary algorithms require a large number of evaluations, which had made the use of these algorithms in real world applications very challenging. This paper presents online evolutionary strategies, as a change of paradigm within CGDA execution. Online evolutionary strategies shift and merge motor execution into the planning loop. A concrete online evolutionary strategy, Online Evolved Trajectories (OET), is presented. OET drastically reduces computational times between motor executions, and enables working in real world dynamic environments and/or with human collaboration. Its performance has been measured against Full Trajectory Evolution (FTE)and Incrementally Evolved Trajectories (IET), obtaining the best overall results. Experimental evaluations are performed on the TEO full-sized humanoid robot with “paint” and “iron” actions that together involve vision, kinesthetic and force features.

We propose a novel approach to establish the relative pose of a mobile robot with respect to an operator that wants to interact with it; we focus on scenarios in which the robot is in the same environment as the operator, and is visible to them. The approach is based on comparing the trajectory of the robot, which is known in the robot's odometry frame, to the motion of the arm of the operator, who, for a short time, keeps pointing at the robot they want to interact with. In multi-robot scenarios, the same approach can be used to simultaneously identify which robot the operator wants to interact with. The main advantage over alternatives is that our system only relies on the robot's odometry, on a wearable inertial measurement unit (IMU), and, crucially, on the operator's own perception. We experimentally show the feasibility of our approach using real-world robots.