ICRA 2015 论文列表

This paper presents a new nonlinear path-following guidance method for autonomous vehicles, which integrates two guidance laws that have given good results independently in applications in autonomous ground, marine, and air vehicles. This new technique retains the best aspects of its two supporting methods, whereas it overcomes some of their drawbacks. It uses the control law by Park et al. to command the vehicle position towards a reference point. However, the position of the reference point is controlled in a different way: instead of being calculated to stay at a fixed distance forward of the vehicle, we use a strategy, inspired in the works by other authors, that controls the speed of the reference point to maintain its position at the given distance. This change does not increment the number of parameters to tune the algorithm and makes it applicable to any initial conditions and parameterized paths. The paper also analyzes the stability of this new nonlinear guidance control law and shows its effectiveness under different simulations.

A novel algorithm is proposed for offline motion planning of a robotic arm to perform a throw task of an object to reach a goal target. The planning algorithm searches for a throw trajectory that could be performed under kinematic and dynamic (i.e., kinodynamic) constraints. We parameterize the throw trajectory by a time-invariant high-dimensional vector. Then, the kinodynamic and target constraints are formulated in terms of time and the parameterization vector. These constraints form time-varying subspaces in the parameterization space. We present a stochastic method for finding a feasible and optimal solution within the subspace. The method generates a number of random points within the parameterization space and checks their feasibility using an adaptive search. The algorithm is guaranteed under a known probability to find a solution if one exists. We present simulations and experiments on a 3R manipulator to validate the method.

A variety of magnetic devices can be manipulated remotely using a single permanent “actuator” magnet positioned in space by a robotic manipulator. This paper presents the spherical-actuator-magnet manipulator (SAMM), which is designed to replace or augment the singularity-prone spherical wrist used by prior permanent-magnet manipulation systems. The SAMM uses three omniwheels to enable holonomic control of a spherical magnet's heading and enable the magnet's instantaneous axis-of-rotation to be set arbitrarily. The SAMM performs closed-loop control of its dipole using field measurements obtained from Hall-effect sensors. We describe the operation and construction of the SAMM, develop and characterize a controller for the SAMM's spherical magnet, and demonstrate remote actuation of an untethered magnetic device in a lumen.

This paper presents the conceptual design, detailed development and flight testing of AtlantikSolar, a 5.6m-wingspan solar-powered Low-Altitude Long-Endurance (LALE) Unmanned Aerial Vehicle (UAV) designed and built at ETH Zurich. The UAV is required to provide perpetual endurance at a geographic latitude of 45°N in a 4-month window centered around June 21st. An improved conceptual design method is presented and applied to maximize the perpetual flight robustness with respect to local meteorological disturbances such as clouds or winds. Airframe, avionics hardware, state estimation and control method development for autonomous flight operations are described. Flight test results include a 12-hour flight relying solely on batteries to replicate night-flight conditions. In addition, we present flight results from Search-And-Rescue field trials where a camera and processing pod were mounted on the aircraft to create high-fidelity 3D-maps of a simulated disaster area.

In this paper, we present the design of a soft wearable exoskeleton that comprises of a glove embedded with pneumatic actuators of variable stiffness for hand assistive and rehabilitation application. The device is lightweight and easily wearable due to the usage of soft pneumatic actuators. A key feature of the device is the variable stiffness of the actuators at different localities that not only conform to the finger profile during actuation, but also provides customizability for different hand dimensions. The actuators can achieve different bending profiles with variable stiffness implemented at different localities. Therefore, the device is able to perform different hand therapy exercises such as full fist, straight fist, hook fist and table top. The device was characterized in terms of its range of motion and maximum force output. Experiments were conducted to examine the differences between active and passive actuation. The results showed that the device could achieve hand grasping and pinching with acceptable range of motion and force.

Substantial progresses have been made in building versatile anthropomorphic prosthetic hands in the past two decades using emerging technologies. However the trade-offs between functionality, reliability, affordability, appearance, etc. have not been fully settled. Many existing designs, particularly the commercial prosthetic hands, are underactuated and they can realize various grasps through compliant structures or differential mechanisms. This paper presents the design of an underactuated prosthetic hand with one actuator using a continuum differential mechanism. Structure of the continuum differential mechanism is simple enough to allow all the components, including a battery pack, to be packed into the palm. The design concept, component descriptions, and hand constructions are elaborated. Experimental verifications are presented to demonstrate the efficacy of the proposed design.

One of the major problems in utilizing origami structures is ensuring variable stiffness; a deployable structure needs to become stiff or flexible according to the requirements of its use in an application. In this study, we present a self-deploying tubular origami mechanism that switches between two distinctive states: small and flexible at its normal state and rigid and stiffened at a locked state. By embedding compact torsional SMA actuators into the mechanism in a novel way through stitching, the process of deploying from the normal to the locked state proceeds in a simple and low-profile manner. With global heating, the torsional SMA wires activate a buckling effect that draws a radical change of folding line from one diagonal to another in every unit tile of the tube, creating axial stiffness. The activated structure, which weighs only 2.9 g, can endure a load of 2.7 kg or more. Additionally, since it does not require bulky actuators, this origami structure can be highly mobile and small in size. This novel origami mechanism is expected to be useful in a wide variety of applications, such as aerospace equipment, mobile architecture, and medical devices, especially those used in minimally invasive surgery (MIS).

In this paper we present a unique self stabilizing ellipsoidal robot for inspection of underwater structures using the principles of ground effect. Underwater metal structures - whether it is ship hulls or internals of a boiling water reactor - require subsurface inspection to detect internal cracks, hidden cavities and other structural damage. This is usually done with on-contact ultrasonic sensors, a slow process if the structure is not sufficiently smooth. However, ultrasound can also be used with a precisely controlled gap. Such precision is challenging relying solely on actuators for control. This paper exploits near surface hydrodynamics to self stabilize a body at a precise gap. Specifically we show how boundary layer and venturi effects combine to create a stable, zero force position at a very small distance from the surface - conceptually similar to air bearings sliders on hard disk drives. Below the stable point lift force dominates, while above it Venturi suction prevails, each bringing the body back to equilibrium. Limitations in the restoring force are considered in the stability analysis included in the paper. This self stabilization method opens a whole new door for non-contact subsurface inspection of underwater structures by autonomous vehicles as well as precision distance maintenance in underwater environment. Here we present our initial analysis and preliminary experimental results for the method when used with an ellipsoidal robot.

The realization and utilization of multimodal locomotion to enable robots to accomplish useful tasks is a significantly challenging problem in robotics. Related to the challenge, it is crucial to notice that the locomotion dynamics of the robots is a result of interactions between a particular control structure and its body-environment dynamics. From this perspective, this paper presents a simple control structure known as Attractor Selection Mechanism that enables a robot to self organize its multiple locomotion modes for accomplishing a goal-directed locomotion task. Despite the simplicity, the approach enables the robot to automatically explore different body-environment dynamics and stabilize onto particular attractors which corresponds to locomotion modes relevant to accomplish the task. The robot used throughout the paper is a curved-beam hopping robot, which despite its simple actuation method, possesses rich and complex body-environment dynamics.

We propose a novel scheme for manipulating a passive object using an active plate. The objective of this study is to control an object's orientation with respect to the gravitational force direction by using an active plate for realizing hitherto unrealized object motion. In this context, motions of the object and active plate are designed to be cyclic. A state vector composed of the object's angle and angular velocity is defined, and the cyclic motion is expressed as a nonlinear discrete system. Fixed points of the state vector are searched for in the designed cyclic motion. A stability analysis around the fixed points is conducted using a Poincaré map. As a result, the fixed points are shown to be asymptotically stable. Finally, experimental results are used to verify that the object's angle can be manipulated with the designed cyclic motion using the plate.

Our goal is to automatically recognize hand grasps and to discover the visual structures (relationships) between hand grasps using wearable cameras. Wearable cameras provide a first-person perspective which enables continuous visual hand grasp analysis of everyday activities. In contrast to previous work focused on manual analysis of first-person videos of hand grasps, we propose a fully automatic vision-based approach for grasp analysis. A set of grasp classifiers are trained for discriminating between different grasp types based on large margin visual predictors. Building on the output of these grasp classifiers, visual structures among hand grasps are learned based on an iterative discriminative clustering procedure. We first evaluated our classifiers on a controlled indoor grasp dataset and then validated the analytic power of our approach on real-world data taken from a machinist. The average F1 score of our grasp classifiers achieves over 0.80 for the indoor grasp dataset. Analysis of real-world video shows that it is possible to automatically learn intuitive visual grasp structures that are consistent with expert-designed grasp taxonomies.

Robotic devices have been shown to be efficacious in the delivery of therapy to treat upper limb motor impairment following stroke. However, the application of this technology to other types of neurological injury has been limited to case studies. In this paper, we present a multi degree of freedom robotic exoskeleton, the MAHI Exo II, intended for rehabilitation of the upper limb following incomplete spinal cord injury (SCI). We present details about the MAHI Exo II and initial findings from a clinical evaluation of the device with eight subjects with incomplete SCI who completed a multi-session training protocol. Clinical assessments show significant gains when comparing pre- and post-training performance in functional tasks. This paper explores a range of robotic measures capturing movement quality and smoothness that may be useful in tracking performance, providing as feedback to the subject, or incorporating into an adaptive training protocol. Advantages and disadvantages of the various investigated measures are discussed with regard to the type of movement segmentation that can be applied to the data collected during unassisted movements where the robot is backdriven and encoder data is recorded for post-processing.

Slip is the major cause of falls in human locomotion. We present a new bipedal modeling approach to capture and predict human walking locomotion with slips. Compared with the existing bipedal models, the proposed slip walking model includes the human foot rolling effects, the existence of the double-stance gait and active ankle joints. One of the major developments is the relaxation of the nonslip assumption that is used in the existing bipedal models. We conduct extensive experiments to optimize the gait profile parameters and to validate the proposed walking model with slips. The experimental results demonstrate that the model successfully predicts the human recovery gaits with slips.

This paper deals with Reconfigurable Cable-Driven Parallel Robots (RCDPRs). A RCDPR is able to change the locations of its cable exit points, the latter being defined as the connection points between the cables and the robot base frame. Given a RCDPR, a set of possible reconfigurations, a desired platform path and a description of the robot environment, the reconfiguration strategy proposed in this paper selects the optimal configurations to be associated to each point of the desired path. The selection of the optimal configurations can be performed with respect to several criteria such as the number of configuration changes, the number of cable reconfigurations and the robot stiffness. In this paper, the optimization is performed using a Dijkstra's based algorithm.

We have developed a contraction and expansion-type artificial muscle by using the plasticized polyvinyl chloride (PVC) gel and mesh electrodes. The PVC gel artificial muscle exhibits a fast response in air, large deformation, variable stiffness, high output force and low power consumption under an electrical field, so it has a great potential to use as a new type of artificial muscles. In a previous study, we designed variable stiffness spats for walking assistance by incorporating the variable stiffness PVC gel artificial muscles with the generally used spats. The spats can assist walking by the generation force of the variation of stiffness with the on and off switching of an electric field. It was found that both the integrated electromyogram (IEMG) and maximal voluntary contraction (%MVC) of the rectus femoris muscle decreased during the walking when wearing the variable stiffness gel spats which showed that it was possible to use the PVC gel artificial muscles for walking assistance. However, the variable stiffness spats has small generation force and lower robustness to external forces. In this study, we propose a novel approach to make a high-performance light-weight walking assist wear using the PVC gel artificial muscles. We use the contraction-expansion output force for motion assist to get a bigger generation force. An expansion type structure unit is introduced to make the assist wear more robust to external forces. Insole force sensors are used to detect the gait changes during walking. The proposed walking assist wear has the characteristics of a simple structure, light weight, easy to put on and take off and with a high flexibility. In this paper, the framework involving the proposed walking assist wear is described in detail.

This paper presents the first method to compute dense scene flow in real-time for RGB-D cameras. It is based on a variational formulation where brightness constancy and geometric consistency are imposed. Accounting for the depth data provided by RGB-D cameras, regularization of the flow field is imposed on the 3D surface (or set of surfaces) of the observed scene instead of on the image plane, leading to more geometrically consistent results. The minimization problem is efficiently solved by a primal-dual algorithm which is implemented on a GPU, achieving a previously unseen temporal performance. Several tests have been conducted to compare our approach with a state-of-the-art work (RGB-D flow) where quantitative and qualitative results are evaluated. Moreover, an additional set of experiments have been carried out to show the applicability of our work to estimate motion in real-time. Results demonstrate the accuracy of our approach, which outperforms the RGB-D flow, and which is able to estimate heterogeneous and non-rigid motions at a high frame rate.

Soft robotics has garnered great research interest from robotics community due to their compliant control as compared to normal hard rigid robots. Particularly, these soft actuators only involve simple design of pneumatic channels to generate required actuation and are usually powered by pneumatics or hydraulics. These actuators are usually lightweight, inexpensive and easily fabricated with soft lithography technique. In this paper, we present a different application of soft robotics in shoe insole design that involves simple design and control. To our knowledge, this is the first study to design soft sensors and actuators which aim to redistribute the plantar pressure and eventually tackle foot problems such as ulcers and blisters. The abnormal, high repetitive stresses acting on the foot tissues can cause immediate breakdown of foot tissues, which will lead to ulcers and even a serious pathology that require foot amputation. A mold with a peripheral pneumatic ring channel feature was designed using computer-aided drawing and thereafter 3D-printed. Subsequently, elastomeric material was poured into the mold and cured to create a negative replica of the mold, which is then sealed using another 2mm layer of elastomeric material. Upon pressurization, the pneumatic channel embedded in the actuator will inflate in the regions that are most compliant and cushion the plantar soft tissues as well as reduce the peak plantar pressure.

Narrow passages in changing environments create huge difficulties, since locations and shapes of narrow passages in Configuration Space(C-space) change frequently. It is very important for a planner to identify narrow passages in real time and boost valid points within them effectively. A novel narrow passage predictive model for designing a path planner in changing environments is proposed in this paper. Firstly, an Expanded Dynamic Bridge Builder is presented to identify narrow passages rapidly with validity-toggle sampling points in C-space. Secondly, the predictive model is adopted to sample possibly free points within these narrow passages without invoking any collision detection in order to avoid intense computational complexity. The predictive model is obtained by the famous classification method of Support Vector Machine (SVM). A new feature, which includes a group of points' distance and validity information, is proposed in SVM training process to capture approximate structure of local narrow passages. Therefore, the predictive model can excavate the hidden similar structure of local narrow passages. Experiments carried out with two 6-DOFs manipulators show that our approach gain higher success rate of planning and time efficiency than other related methods.

We present a novel approach for generating candidate handover positions for human receivers. The object handover problem involves several variables and is under-constrained. The sheer number of possibilities makes it nontrivial to find a good object handover position for existing algorithms. To address this problem, we introduce the use of a custom biomechanical model in a novel algorithm for generating a handover position. The biomechanical model serves as a tool for approximating the static strength a human receiver needs to exert in order to receive an object. A mobile manipulator can use the output of our algorithm when calculating a target position for handing over an object to a human. We show preliminary results involving electromyography data and joint moment values, as well as a proof-of-concept implementation with a mobile manipulator to fetch a toolbox to a receiver.