ICRA 2018 论文列表

This paper studies dexterous manipulation in the plane by a two-fingered hand in the plane. The dynamics of each finger, which consists of two links with coupled joints, are derived based on Lagrangian mechanics. As an object is being manipulated, its orientation and the two independent joint angles of the hand constitute the state of the entire system. Contact kinematics, accounting for both stick and slip modes, are combined with dynamics to establish a dependence of the object's linear and angular accelerations on joint accelerations. This allows control of joint torques, under a proportional-derivative (PD) law, to move the object to a target position in a desired orientation.

Vacuum-based end effectors are widely used in industry and are often preferred over parallel-jaw and multifinger grippers due to their ability to lift objects with a single point of contact. Suction grasp planners often target planar surfaces on point clouds near the estimated centroid of an object. In this paper, we propose a compliant suction contact model that computes the quality of the seal between the suction cup and local target surface and a measure of the ability of the suction grasp to resist an external gravity wrench. To characterize grasps, we estimate robustness to perturbations in end-effector and object pose, material properties, and external wrenches. We analyze grasps across 1,500 3D object models to generate Dex-Net 3.0, a dataset of 2.8 million point clouds, suction grasps, and grasp robustness labels. We use Dex-Net 3.0 to train a Grasp Quality Convolutional Neural Network (GQ-CNN) to classify robust suction targets in point clouds containing a single object. We evaluate the resulting system in 350 physical trials on an ABB YuMi fitted with a pneumatic suction gripper. When evaluated on novel objects that we categorize as Basic (prismatic or cylindrical), Typical (more complex geometry), and Adversarial (with few available suction-grasp points) Dex-Net 3.0 achieves success rates of 98%, 82%, and 58% respectively, improving to 81% in the latter case when the training set includes only adversarial objects. Code, datasets, and supplemental material can be found at http://berkeleyautomation.github.io/dex-net.

To achieve immersive and natural navigation in a virtual environment through human locomotion, it is necessary to generate a 2-dimensional infinite ground for omnidirectional walking. However, the existing omnidirectional treadmills are heavy, complex and exhibit low acceleration due to power transmission inefficiency. In this paper, we present a novel fast-omnidirectional treadmill (F-ODT) with a new power transmission mechanism called the Geared Omni-pulley. This mechanism ensures higher power transmission efficiency for driving the belts of the multiple transversal treadmills for independent Y-axis motion. Due to the improved power transmission performance combined with a simpler and relatively light-weight structure, the proposed 2D treadmill can generate a maximum speed of 3m/sec with an acceleration of 3m/sec2. Based on the improved performance, the F-ODT system can be used as a locomotion interface platform in various virtual reality environments such as training of soldiers, gaming/educational experiences and gait rehabilitation.

This paper presents the design, development and implementation of a novel, high power hexapole magnetic tweezer system for 3D micromanipulations. Six tapering-tipped magnetic poles are deployed in a tilted Cartesian coordinate system, with an electromagnetic coil on each for actuation, connected by two 3D printed magnetic yokes to form a double layer structure. The power source is integrated to the magnetic tweezer system through a control algorithm on the software level; image processing was used for experiment analysis. Because of the high magnetic field that the magnetic coils can generate, the working space in the system is relatively larger than other similar designs, which provides better performance on microscale robotic swimmer manipulations. Simulations and experiments performed in this paper demonstrate the agile and powerful manipulation of microswimmers with desired control input to follow complex trajectories, avoid obstacles and move against micro-flow in the samples. We prove that the developed hexapole magnetic tweezer has enough power and controllability to guide microswimmers in Newtonian and Non-Newtonian fluid environments. The system will be optimized continuously and implemented into cell penetration experiments. Finally, the application will be deployed into in vivo based environments.

Retinal Vein Occlusion is a common retinal vascular disorder which can cause severe loss of vision. Retinal vein cannulation followed by injection of an anti-coagulant into the affected vein is a promising treatment. However, given the scale and fragility of the surgical workfield, this procedure is considered too high-risk to perform manually. A first successful robot-assisted procedure has been demonstrated. Even though successful, the procedure remains extremely challenging. This paper aims at providing a solution for the limited perception of instrument-tissue interaction forces as well as depth estimation during retinal vein cannulation. The development of a novel combined force and distance sensing cannulation needle relying on Fiber Bragg grating (FBG) and Optical Coherence Tomography (OCT) A-scan technology is reported. The design, the manufacturing process, the calibration method, and the experimental characterization of the produced sensor are discussed. The functionality of the combined sensing modalities and the real-time distance estimation algorithm are validated respectively on in-vitro and ex-vivo models.

In this paper, we present a novel method to detect and resolve motion conflicts in visual-inertial odometry. Recently, it has been common to integrate an IMU sensor with visual odometry in order to improve localization accuracy and robustness. However, when a disagreement between the two sensor modalities occurs, the localization accuracy reduces drastically and leads to irreversible errors. In such conditions, multiple motion estimates based on the set of observations used are possible. This creates a conflict (motion conflict) in determining which observations to use for accurate ego-motion estimation. Therefore, we present a method to detect motion conflicts based on per-frame positional estimate discrepancy and per-landmark reprojection errors. Additionally, we also present a method to resolve motion conflicts by eliminating inconsistent IMU and landmark measurements. Finally, we implement Motion Conflict aware Visual Inertial Odometry (MC-VIO) by combining both detection and resolution of motion conflicts. We perform quantitative and qualitative evaluation of MC-VIO on visually and inertially challenging datasets. Experimental results indicate that the MC-VIO algorithm reduces the increase in absolute trajectory error by 80% and the relative pose error by 60% for scenes with motion conflict, in comparison to the state-of-the-art reference VIO algorithm.

In this research, it is shown that robot joint torques can be used to recognize contact force transients induced during robotic manipulation, thus detecting when a task is completed. The approach does not assume any external sensor, which is a benefit compared to the state of the art. The joint torque data are used as input to a recurrent neural network (RNN), and the output of the RNN indicates whether the task is completed. A real-time application for force transient detection is developed, and verified experimentally on an industrial robot.

This paper presents a new design of multirotor, named as `Tilting Thruster Type' (T3)-multirotor. The new platform is equipped with mechanically separated thrusters, which can take any fuselage posture within a specified range regardless of any direction of translational acceleration. A specially designed servo-linkage mechanism is employed for relative attitude control between the thruster and the fuselage. Mathematical modeling and analysis of the new platform are conducted to explore the control method of the dynamically complex system. For demonstrating the potential of the new T3-multirotor, an autonomous level flight is performed where the fuselage maintains zero roll and pitch angle during the entire flight. Both simulation and experimental results are provided with detailed analysis.

In this paper, we present the design, modeling, and real-time nonlinear model predictive control (NMPC) of an autonomous robotic boat. The robot is easy to manufacture, highly maneuverable, and capable of accurate trajectory tracking in both indoor and outdoor environments. In particular, a cross type four-thruster configuration is proposed for the robotic boat to produce efficient holonomic motions. The robot prototype is rapidly 3D-printed and then sealed by adhering several layers of fiberglass. To achieve accurate tracking control, we formulate an NMPC strategy for the four-control-input boat with control input constraints, where the nonlinear dynamic model includes a Coriolis and centripetal matrix, the hydrodynamic added mass, and damping. By integrating “GPS” modules and an inertial measurement unit (IMU) into the robot, we demonstrate accurate trajectory tracking of the robotic boat along preplanned paths in both a swimming pool and a natural river. Furthermore, the code generation strategy employed in our paper yields a two order of magnitude improvement in the run time of the NMPC algorithm compared to similar systems. The robot is designed to form the basis for surface swarm robotics testbeds, on which collective algorithms for surface transportation and self-assembly of dynamic floating infrastructures can be assessed.

This paper presents the design, modeling, analysis, and experimental validation of an inductive resonant wireless power transfer (WPT) system to power a micro aerial vehicle (MAV). Using WPT, in general, enables longer flight times, virtually eliminates the need for batteries, and minimizes down time for recharging or replacing batteries. The proposed WPT system consists of a transmit coil, which can either be fixed to ground or placed on a mobile platform, and a receive coil carried by the MAV. The details of the WPT circuit design are presented. A power-transfer model is developed for the two-coil system, where the model is used to select suitable coil geometries to maximize the power received by the MAV for hovering. Analysis, simulation, and experimental results are presented to demonstrate the effectiveness of the WPT circuitry. Finally, a wirelessly powered MAV that hovers above the transmit coil is demonstrated in a laboratory setting.

This paper presents the design, modeling, control, and experimental test of a solar-powered quadcopter to allow for long-endurance missions. We first present the design of a large-scale quadcopter that incorporates solar energy harvesting capabilities. Based on the design results, we built the dynamical model of the customized quadcopter with analysis of the aerodynamic influence. A feedback control system is developed for the solar-powered quadcopter that takes into account the wind disturbance and is verified in virtual simulation examples. All parameters used in the modeling and simulations are based on a developed prototype of the solar-powered quadcopter. Flight tests with the prototype are presented to validate the feasibility and theoretical basis of the solar-powered quadcopter.

In most cases of peripheral arterial disease (PAD), the operating surgeon must use a variety of catheters riding on a thin wire known as a `guidewire'. This guidewire must be manually navigated through a tortuous pathway of arteries to arrive at the diseased area. Automation of the guidewire therefore reduces surgeon effort and minimizes the time required for a PAD procedure, but is restricted by the size constraints of a standard guidewire. This work presents the design of a robotically actuated 2 degree-of-freedom (DoF) guidewire tip comprised of joints laser micro-machined into a 0.78 mm (<; 2.4 Fr) Nitinol tube. We present an analysis of the notch joint used as a building block in the robot and a control strategy for this type of a joint. The experimental results show that tendon force is an important observable quantity that can be used as a shape sensing mechanism for this type of a joint in practical control applications.

This paper introduces jlüela driverless: the first autonomous racecar to win a Formula Student Driverless competition. In this competition, among other challenges, an autonomous racecar is tasked to complete 10 laps of a previously unknown racetrack as fast as possible and using only onboard sensing and computing. The key components of flüela's design are its modular redundant sub-systems that allow robust performance despite challenging perceptual conditions or partial system failures. The paper presents the integration of key components of our autonomous racecar, i.e., system design, EKF-based state estimation, LiDAR-based perception, and particle filter-based SLAM. We perform an extensive experimental evaluation on real-world data, demonstrating the system's effectiveness by outperforming the next-best ranking team by almost half the time required to finish a lap. The autonomous racecar reaches lateral and longitudinal accelerations comparable to those achieved by experienced human drivers.

This paper presents a 6 DOF leg mechanism for a humanoid robot. The proposed Hybrid Leg is designed to combine serial and parallel mechanisms and consists of a pair of twin 3 DOF serial chains in parallel. A 5-bar-linkage mechanism is implemented to the serial mechanism to generate 2 DOF motion regarding hip and knee pitch rotation. The hardware prototype is designed by matching the kinematic specification of a commercial robot's leg to compare the proposed mechanism with a conventional serial leg. We derive the analytical expressions of its forward and inverse kinematics. End-effector workspaces are shown with plots and inverse dynamics analysis of Hybrid Leg and serial leg with a given walking gait trajectory is presented. Hardware experiment is conducted with a prototype to verify the simulated workspace and trajectory tracking performance.

This work presents the development of a 3-DoF leg with series and parallel compliant actuation. Series-elastic main actuators are combined with parallel high efficiency energy storage branches, to substantially improve energy efficiency. The leg design is semi-anthropomorphic, with similar mass and mass distribution to the human limb, and includes a biarticulated actuation configuration. The parallel branches are driven by secondary motors and their design parameters are optimised. The mechanical design of the prototype leg is presented, introducing details of the actuation configuration principles employed. Preliminary experimental data are presented, in which a baseline series-elastic-only configuration is compared with configurations with mono- and biarticulated parallel branches, respectively. The results effectively demonstrate the concept's potential, showing improvements of 53% and 60% in electrical power consumption while the leg is executing loaded cyclic motion profiles.

Soft-bodied and continuum robots have shown great adaptability to the environment thanks to its flexibility of the body. They have great potential in environment exploring or rescuing mission. One of those robots is snake-like soft-bodied robots. A snake robot is often made by attaching passive wheels along a long body to achieve frictional anisotropy. This anisotropic structure helps to propel the body with serpentine locomotion and prevents it from sliding laterally. However, with a snake-like soft-bodied robot, attaching wheels is not only clumsy but also adding weight to the robot. In this paper, being inspired by the scales on the skin of a snake, we propose a designing scheme to achieve an all-printed wriggle soft-bodied robot by patterning high and low friction material to the ventral side of the robot. Compared to a totally flat ventral, we are able to speed-up the serpentine locomotion 2.8 times. Besides, by changing the configuration of high/low friction material, our wriggle soft-bodied robot can easily move forward or backward just by switching the controlling signal. The fabrication time is just less than 1 hour and the robot can achieve the speed of 26 mm/s.

Bronchoscopic intervention, as a minimally invasive method for the diagnosis and treatment of lung diseases, has attracted more and more attention in recent years. However, existing endobronchial instruments lack the steerability accessing the peripheral airways with difficult bifurcations. This paper presents a novel wire-driven dexterous manipulator for the guidance of such instruments. Precision laser profiling is used to cut a stainless steel tube into multiple interlocked segments with revolute joints. The outer diameter of the manipulator is 2.20 mm which is small enough to be inserted into the working channels of most commercial bronchoscopes and distal airways, while keeping a large inner lumen with a diameter of 1.44 mm for passing various bronchoscopic instruments. The small bending radius provides enough flexibility to navigate inside the complex bronchial tree. Two kinematic models are proposed to predict the manipulator configuration from the translation of actuation wires. The former model is geometrically derived with the assumption of constant curvature bending and the latter one is statistically driven by capturing the motion trajectories of manipulator joints. A prototype of our low-cost add-on instrument guidance robot for bronchoscopic intervention is presented which can be easily integrated into current clinical routine.

This paper proposes a novel in-vivo robotic laparo-scopic camera design with an optimized illumination system, which is a crucial component for achieving high imaging quality. The robotic camera design with three extendable wings can reserve sufficient on-board space to harbor the optimized illumination system without affecting the compactness of the camera. We contribute a freeform optical lens design method and develop three miniature optical lenses for the LEDs to achieve greater than 95% illumination uniformity, greater than 14, 000 lx illuminance on a target plane with a distance of 100 mm, and greater than 89% optical efficiency. The prototype is implemented and experimentally tested, which demonstrates great performance of the in-vivo robotic laparoscopic camera and the significance of the optimized illumination system.

Microgrippers play a significant role in manipulation of micro-objects. To achieve dexterous and precise manipulation, a microgripper is required to be compactly designed and embedded with sensing feedback. Meanwhile, to convert the sensor position into displacement of the microgripper, the embedded sensors should be calibrated by additional equipment like laser sensor. However, a microgripper always needs to be calibrated during manipulation (online calibration), which is still a big challenge with current technology. In this paper, we proposed a highly compact microgripper integrated with position sensors, and a visual-based calibration method to handle such challenge. Moreover, to enhance grasping accuracy, flexure hinges are employed to achieve a low impedance grasping mechanism and to avoid the backlash in traditional bearing. Furthermore, kinematics analysis and Fine Element Analysis (FEA) are implemented to improve the design efficiency. Finally, fibrous micro-rings are successfully assembled, and the results reveal that the calibrated microgripper can be well employed to operate micro-objects.