ICRA 2018 论文列表

We present a collaborative visual localization method for rovers designed to hop and tumble across the surface of small Solar System bodies, such as comets and asteroids. In a two-phase approach, an orbiting primary spacecraft first maps the surface of a body by capturing images from various poses and illumination angles; these images are processed to create a prior map of 3D landmarks. In the second phase, a hopping rover is deployed to the surface where it uses a camera to relocalize to the prior map and to perform onboard visual simultaneous localization and mapping (SLAM). Small bodies present several unique challenges to existing visual SLAM algorithms, such as high-contrast shadows that move quickly over the surface due to the short (e.g. 1-12 hour) rotational periods, and large changes in visual appearance between orbit and the surface, where image scale varies by many orders of magnitude (kilometers to centimeters). In this work, we describe how to augment ORB-SLAM2-a state of the art visual SLAM implementation-to handle large variations in illumination by fusing prior images with varying illumination angles. We demonstrate how a hopping rover can use a wide field of view (FOV) camera to relocalize to prior maps captured by an orbiting spacecraft with a narrow FOV camera, and how the growth of pose and scale errors can be bounded by periodic loop closures during large hops. The proposed method is evaluated with sequences of images captured around a mock asteroid; it is shown to be robust to varying illumination angles, scene scale changes, and off-nadir camera pointing angles.

In this paper we propose a method to achieve relative positioning and tracking of a target by a quadcopter using Ultra-wideband (UWB) ranging sensors, which are strategically installed to help retrieve both relative position and bearing between the quadcopter and target. To achieve robust localization for autonomous flight even with uncertainty in the speed of the target, two main features are developed. First, an estimator based on Extended Kalman Filter (EKF) is developed to fuse UWB ranging measurements with data from onboard sensors including inertial measurement unit (IMU), altimeters and optical flow. Second, to properly handle the coupling of the target's orientation with the range measurements, UWB based communication capability is utilized to transfer the target's orientation to the quadcopter. Experiments results demonstrate the ability of the quadcopter to control its position relative to the target autonomously in both cases when the target is static and moving.

Robots working in natural, urban, and industrial settings need to be able to navigate challenging environments. In this paper, we present a motion planner for the perceptive rough-terrain locomotion with quadrupedal robots. The planner finds safe footholds along with collision-free swing-leg motions by leveraging an acquired terrain map. To this end, we present a novel pose optimization approach that enables the robot to climb over significant obstacles. We experimentally validate our approach with the quadrupedal robot ANYmal by autonomously traversing obstacles such steps, inclines, and stairs. The locomotion planner re-plans the motion at every step to cope with disturbances and dynamic environments. The robot has no prior knowledge of the scene, and all mapping, state estimation, control, and planning is performed in real-time onboard the robot.

Fast and robust person detection is one of the most important tasks for robotic applications involving human interaction. Particularly in mobile robotics this task is still challenging. Though there are already reliable and real-time capable approaches, they are usually computationally expensive. They either require GPUs or multiple CPU cores in order to work properly. Furthermore, some of the approaches are designed for special environments and sensor types, which reduces general applicability. In this work, we present a robust, generic and lightweight solution for real-time 3D person detection. Since our approach requires only a single CPU thread, it can be run as a background process and is suitable for smaller robotic systems. We demonstrate applicability to indoor and outdoor scenarios using different 3D sensor types separately. Moreover, we are able to show that the proposed method outperforms other state-of-the-art approaches, including a DCNN.

This paper presents a manifold based Unscented Kalman Filter that applies a novel strategy for inertial, model-aiding and Acoustic Doppler Current Profiler (ADCP) measurement incorporation. The filter is capable of observing and utilizing the Earth rotation for heading estimation with a tactical grade IMU, and utilizes information from the vehicle model during DVL drop outs. The drag and thrust model-aiding accounts for the correlated nature of vehicle model parameter error by applying them as states in the filter. ADCP-aiding provides further information for the model-aiding in the case of DVL bottom-lock loss. Additionally this work was implemented using the MTK and ROCK framework in C++, and is capable of running in real-time on computing available on the FlatFish AUV. The IMU biases are estimated in a fully coupled approach in the navigation filter. Heading convergence is shown on a real-world data set. Further experiments show that the filter is capable of consistent positioning, and data denial validates the method for DVL dropouts due to very low or high altitude scenarios.

In this study, we propose a novel Light Detection and Ranging (LiDAR) sensor-based localization method for localization of a mobile robot. In localization using the LiDAR sensor, localization errors occur when real range measurements differ from reference distances computed from a map. This study focuses on three factors that cause differences between real range measurements and reference distances. The first factor corresponds to optical characteristics of the LiDAR sensor for objects such as glass walls and mirrors. The second factor corresponds to occlusions by dynamic obstacles. The third factor corresponds to static changes in the environment. In practical applications, three factors often simultaneously occur. Although there have been many previous works, robust localization to overcome these three difficulties is still a challenging problem. This study proposes a novel robust localization scheme that exploits only reliable range measurements. A LiDAR sensor-based localization scheme can be successfully executed by utilizing only a few reliable range measurements. Therefore, the computation of reliability plays a significant role. The computation of reliability is divided into two steps. The first step considers characteristics of optical sensors. The second step mainly deals with the effects of obstacles. The observation likelihood model exploits computed reliability for pose estimation. The proposed scheme was successfully verified through various experiments under challenging situations.

The capability of following a person is crucial in service-oriented robots for human assistance and cooperation. Though a vast variety of following systems exist, they lack robustness against dynamic changes of the environment and relocating to continue following a lost target. Here we present a robust human following system that has the extendability to commercial service robot platforms having a RGB-D camera. The proposed framework integrates deep learning methods for perception and variational Bayesian techniques for trajectory prediction. Deep learning modules enable robots to accompany a person by detecting the target, learning the target and following while avoiding collision within the dynamic home environment. The variational Bayesian techniques robustly predict the trajectory of the target by empowering the following ability of the robot when target is lost. We experimentally demonstrate the capability of the deep Bayesian trajectory prediction method on real-time usage, following abilities, collision avoidance and trajectory prediction of the system. The proposed system was deployed at the RoboCup@Home 2017 Social Standard Platform League and successfully demonstrated its robust functions and smooth person following capability resulting in winning the 1st place.

This paper introduces a robust point cloud registration method which utilizes not only positional but also the orientation information at each point. The proposed method takes a probabilistic approach which forms the problem as a hybrid mixture model, in which a Von-Mises-Fisher mixture model (FMM) is adopted to model the orientation part and a gaussian mixture model (GMM) is used to represent the position part. When two point clouds are optimally registered, the correspondence is the maximum of the posterior probability of the overall mixture model. Expectation-Maximization (EM) algorithm has been adopted to solve the optimization problem in an iterative manner to find the optimal rotation and translation between two point clouds. Extensive experiments under different noise levels and different outlier ratios have been carried out on a dataset of the femur CT images. Comparison results show that the proposed method outperforms the state-of-the-art methods under most of the experimental conditions, which indicates the validity of our method.

Constructing a spatial map of environmental parameters is a crucial step to preventing hazardous chemical leakages, forest fires, or while estimating a spatially distributed physical quantities such as terrain elevation. Although prior methods can do such mapping tasks efficiently via dispatching a group of autonomous agents, they are unable to ensure satisfactory convergence to the underlying ground truth distribution in a decentralized manner when any of the agents fail. Since the types of agents utilized to perform such mapping are typically inexpensive and prone to failure, this results in poor overall mapping performance in real-world applications, which can in certain cases endanger human safety. This paper presents a Bayesian approach for robust spatial mapping of environmental parameters by deploying a group of mobile robots capable of ad-hoc communication equipped with short-range sensors in the presence of hardware failures. Our approach first utilizes a variant of the Voronoi diagram to partition the region to be mapped into disjoint regions that are each associated with at least one robot. These robots are then deployed in a decentralized manner to maximize the likelihood that at least one robot detects every target in their associated region despite a non-zero probability of failure. A suite of simulation results is presented to demonstrate the effectiveness and robustness of the proposed method when compared to existing techniques.

We present a stereo-based dense mapping algorithm for large-scale dynamic urban environments. In contrast to other existing methods, we simultaneously reconstruct the static background, the moving objects, and the potentially moving but currently stationary objects separately, which is desirable for high-level mobile robotic tasks such as path planning in crowded environments. We use both instance-aware semantic segmentation and sparse scene flow to classify objects as either background, moving, or potentially moving, thereby ensuring that the system is able to model objects with the potential to transition from static to dynamic, such as parked cars. Given camera poses estimated from visual odometry, both the background and the (potentially) moving objects are reconstructed separately by fusing the depth maps computed from the stereo input. In addition to visual odometry, sparse scene flow is also used to estimate the 3D motions of the detected moving objects, in order to reconstruct them accurately. A map pruning technique is further developed to improve reconstruction accuracy and reduce memory consumption, leading to increased scalability. We evaluate our system thoroughly on the well-known KITTI dataset. Our system is capable of running on a PC at approximately 2.5Hz, with the primary bottleneck being the instance-aware semantic segmentation, which is a limitation we hope to address in future work. The source code is available from the project websiteaahttp://andreibarsan.github.io/dynslam.

The control of walking robots has been a long-studied problem as researchers attempt to bring robots out of the lab and into unstructured environments. In particular, there is significant interest in dynamic walkers: a class of walking robots that exhibit highly efficient gaits, but are sensitive to disturbances. This paper develops robust controllers for dynamic walkers in transverse coordinates using H∞ control, with a focus on rejecting disturbances at foot impact. The optimization objective is a measure of disturbance rejection known as the gait sensitivity norm. The controller was used to stabilise a compass gait walker in simulation for various disturbances. Simulation results demonstrate the advantages of phase-tracking and H∞ control for robustness compared to time-varying and LQR controllers.

Recent advances in perception and planning algorithms have enabled robots to navigate autonomously through unknown, cluttered environments at high-speeds. A key component of these systems is the ability to identify, select, and execute a safe trajectory around obstacles. Many of these systems, however, lack performance guarantees because model uncertainty and external disturbances are ignored when a trajectory is selected for execution. This work leverages results from nonlinear control theory to establish a bound on tracking performance that can be used to select a provably safe trajectory. The Composite Adaptive Sliding Controller (CASC) provides robustness to disturbances and reduces model uncertainty through high-rate parameter estimation. CASC is demonstrated in simulation and hardware to significantly improve the performance of a quadrotor navigating through unknown environments with external disturbances and unknown model parameters.

This paper introduces an approach that robotizes an ankle-foot orthosis (AFO). In particular, toward post-stroke gait rehabilitation, we robotize a double-bar AFO, which is widely used in rehabilitation facilities, by newly designing a modular joint, a pneumatic actuator, and a Bowden cable force-transmission system. Our modular joint system, called the Modular Exoskeletal Joint (MEJ), has a hollow shaft for simple attachment to an AFO's pivot. We designed MEJ to compactly house an encoder that is built in a bearing in a pulley. We adopted Bowden cables to transmit contraction forces from an actuator to the MEJ. As an actuation scheme, we developed the Nested-cylinder Pneumatic Artificial Muscle (NcPAM) system. Even though PAMs are mechanically compliant and lightweight, they can still generate a large force. Therefore, they can provide an ideal actuation system for exoskeletal robots. The nested-cylinder in NcPAM houses a cable-tensioning spring to properly maintain small cable tension for passive movements and a cable stopper to connect the PAM and the cable for properly transmitting the large force generated by PAM. We show the ankle-joint trajectory tracking performances of this integrated system using iterative learning control.

This paper presents a robotic pick-and-place system that is capable of grasping and recognizing both known and novel objects in cluttered environments. The key new feature of the system is that it handles a wide range of object categories without needing any task-specific training data for novel objects. To achieve this, it first uses a category-agnostic affordance prediction algorithm to select and execute among four different grasping primitive behaviors. It then recognizes picked objects with a cross-domain image classification framework that matches observed images to product images. Since product images are readily available for a wide range of objects (e.g., from the web), the system works out-of-the-box for novel objects without requiring any additional training data. Exhaustive experimental results demonstrate that our multi-affordance grasping achieves high success rates for a wide variety of objects in clutter, and our recognition algorithm achieves high accuracy for both known and novel grasped objects. The approach was part of the MIT-Princeton Team system that took 1st place in the stowing task at the 2017 Amazon Robotics Challenge. All code, datasets, and pre-trained models are available online at http://arc.cs.princeton.edu.

Magnetic micromanipulation is an untethered technique and has enabled numerous applications in the scale of millimeters to micrometers from the tissue level to cell level. However, existing systems are not capable of maneuvering a sub-micrometer object for precise force control, preventing the realization of intracellular manipulation or `fantastic voyage' inside a single cell. The magnetic micromanipulation task achieved in this work is sub-micrometer position control and piconewton force control of a sub-micron (0.7 μm) magnetic bead inside a single human bladder cancer cell (RT4). The magnetic bead was 3D positioned in the cell using a generalized predictive controller that effectively tackled the control challenge caused by the slow visual feedback (1 Hz) from high-resolution confocal microscopy. The average positioning error was quantified to be 0.43 μm, which is slightly larger than Brownian motion-imposed constraint (0.31 μm). The system is capable of three-dimensionally applying a maximum force of 60 pN with a resolution of 4 pN. In experiments, a 0.7 μm magnetic bead was controlled to move from an initial position in a cell to target positions on the cell nucleus. Force-displacement data were obtained from multiple locations along the cell nucleus' major and minor axes. The results revealed, for the first time, significantly higher stiffness exists in the cell nucleus' major axis than the minor axis. This stiffness polarity was likely attributed to the aligned stress fibers of actin filament inside the cells.

Manipulation of motile cells such as bacteria and sperm is required in both cell biology and clinical applications. For immobilizing a motile sperm, the sperm head and tail positions must be accurately tracked, interference of proximal sperms on the target sperm must be tackled, and the orientation of the sperm must be properly aligned with the manipulation tool in order not to damage the sperm head where DNA is contained. Manual operation of sperm immobilization has stringent skill requirements, and both manual operation and existing robotic sperm immobilization suffer from inconsistent success rates and incapability of manipulating sperms swimming in all directions. This paper presents a robotic system for fully automated tracking, orientation control, and immobilization of motile sperms. Algorithms were developed for robustly tracking the sperm head and estimating the sperm tail positions under interfering conditions. A new visual servo control strategy was developed to enable the robotic system to actively adjust sperm orientation for immobilizing a sperm swimming in any direction. Experimental results from robotic immobilization of 400 sperms confirmed that the robotic system achieved a consistent success rate of 94.5 %, independent of sperm velocity or swimming direction.

Handling liquids with spilling avoidance is a topic of interest for a broad range of fields, both in industry and in service robotic applications. In this paper we present a new control architecture for motion planning of industrial robots, able to tackle the problem of liquid transfer with sloshing control. We do not focus on a complete sloshing suppression, but we show how to enforce an anti spilling constraint. This less conservative approach allows to impose higher accelerations, reducing motion time. A constraint-based approach, amenable to an Online implementation, has been developed. The proposed controller generates trajectories in real time, in order to follow a reference path, while being compliant to the spilling avoidance constraint. The approach has been validated on a 6 degree of freedom industrial ABB robot.

We address the problem of enabling a manipulator to move arbitrary amounts and configurations of dirt on a surface to a goal region using a cleaning tool. We represent this problem as heuristic search with a set of primitive dirt-oriented tool actions. We present dirt and action representations that allow efficient learning and prediction of future dirt states, given the current dirt state and applied action. We also present a method for sampling promising actions based on a clustering of dirt states and heuristics for planning. We demonstrate the effectiveness of our approach on challenging cleaning tasks through implementations on PR2 and Fetch robots.

Hands-on training is an indispensable part of surgical practice. As the tools used in the operating room become more intricate, the demand for efficient training methods increases. This work proposes a robotic assistance-as-needed method for training with surgical teleoperated robots. The method adapts the intensity of the assistance according to the trainee's current and past performance while gradually increasing the level of control of the trainee as the training progresses. The work includes an experiment comprising 160 acquisition sessions from 16 novice subjects performing a bimanual teleoperated exercise with a da Vinci Research Kit surgical console. Results capture the subtleties in the task's learning curve with and without robotic assistance and hint at the potential of robotic assistance for complex visuomotor training. Although robotic assistance for motor learning has received mixed results that range from beneficial to detrimental effects, this study shows such assistance may increase the rate of learning of certain skills in complex motor tasks.