Shinichi Hirai

Robust real time material classification algorithm using soft three axis tactile sensor: Evaluation of the algorithm

Materials and textures identification is a desired ability for robots. Developing such systems require tactile sensors that have enough sensitivity and spatial resolution, and the computational intelligence to meaningfully interpret sensor data. This paper introduces a texture classification algorithm utilizing support vector machine (SVM) classifier. Data taken from a novel three axis tactile sensor that utilize magnetic flux measurements for transduction was used to obtain the three dimensional tactile data. Frobenius norm calculated from the covariance matrix of the above data and the mean values of the three dimensional sensor data were used as features. Palpation velocity and small vertical load variances had minimum influence on the proposed algorithm. We have compared this algorithm with two other classification methods. They are: classify using the feature spatial period that is calculated from principal frequencies of the textures/material, and classify using neural network classifier with special properties of each materials tactile signals as features. For eight classes of material, the proposed algorithm performed faster and more accurately than the comparators when the scanning velocity and the vertical load varied.

Crawling and Jumping by a Deformable Robot

We describe crawling and jumping by a soft robot. Locomotion over rough terrain has been achieved mainly by rigid body systems including crawlers and leg mechanisms. This paper presents an alternative method, one that employs deformation. First, we describe the principle of crawling and jumping as performed through deformation of a robot body. Second, in a physical simulation, we investigate the feasibility of the approach. Next, we show experimentally that a prototype of a circular soft robot can crawl and jump. Finally, we describe crawling and jumping performed by a spherical deformable robot.

Kinematics and Statics of Manipulation Using the Theory of Polyhedral Convex Cones

A new approach to the kinematic and static analysis of ma nipulative tasks performed through mechanical contacts is presented. A variety of manipulation problems, including as sembly and grasps, have been treated separately in robotics research. All of the problems are treated as ways to solve a certain class of inequalities resulting from the unidirectional nature of mechanical contacts. One of the fundamental diffi culties in the analysis of manipulative tasks is the intractable nature of inequalities. In this article, we establish an underpinning mathematical tool for dealing with a variety of manipulative tasks that are governed by unidirectional constraints. First, we introduce a coherent representation for formulating various manipulation problems. Second, we develop several procedures based on the theory of polyhedral convex cones to solve these problems in a systematic and straightforward manner. The method is then implemented on a computer and applied to a variety of manipulation problems, including grasping, fixturing, and hybrid position/force control.

Static Modeling of Linear Object Deformation Based on Differential Geometry

We describe the modeling of linear object deformation based on differential geometry and its applications to manipulative operations. A particle-based approach, the finite element method, and the Cosserat theory have been applied to the modeling of linear object deformation. In this paper, we establish an alternative modeling approach based on an extension of differential geometry. First, we extend differential geometry to describe linear object deformation including flexure, torsion, and extension. Secondly, we show computational results to demonstrate the feasibility of the proposed modeling technique, and we compare computational and experimental results to demonstrate the accuracy of the model. Next, we apply the proposed approach to the grasping of a deformable linear object. We propose a disturbance force margin to indicate the stability of the grasping and we describe the computation of the margin using the proposed approach. Finally, we apply the proposed approach to the deformation path planning of a linear object. We formulate the minimization of potential energy during a deformation path. We compute the optimal deformation path and a feasible deformation path, which are compared with an experimental result.

Circular/Spherical Robots for Crawling and Jumping

We describe circular/spherical robots for crawling and jumping. Locomotion over rough terrain has been achieved mainly by rigid body systems including crawlers and leg mechanisms. This paper presents an alternative method of moving over rough terrain, one that employs deformation. First, we describe the principle of crawling and jumping as performed through deformation of a robot body. Second, in a physical simulation, we investigate the feasibility of the approach. Next, we show experimentally that prototypes of a circular robot and a spherical robot can crawl and jump.

Elastic Model of Deformable Fingertip for Soft-Fingered Manipulation

We propose a straightforward static elastic model of a hemispherical soft fingertip undergoing large contact deformation, as occurs when robotic hands with the fingertips handle and manipulate objects, which is suitable for the analysis of soft-fingered manipulation because of the simple form of the model. We focus on formulating elastic force and potential energy equations for the deformation of the fingers which are represented as an infinite number of virtual springs standing vertically. The equations are functions of two variables: the maximum displacement of the hemispherical fingertip and the orientation angle of a contacting planar object. The elastic potential energy has a local minimum in our model. The elastic model was validated by comparison with results of a compression test of the hemispherical soft fingertip

Knotting/Unknotting Manipulation of Deformable Linear Objects

Here, we propose a planning method for knotting/unknotting of deformable linear objects. First, we propose a topological description of the state of a linear object. Second, transitions between these states are defined by introducing four basic operations. Then, possible sequences of crossing state transitions, i.e. possible manipulation processes, can be generated once the initial and the objective states are given. Third, a method for determining grasping points and their directions of movement is proposed to realize derived manipulation processes. Our proposed method indicated that it is theoretically possible for any knotting manipulation of a linear object placed on a table to be realized by a one-handed robot with three translational DOF and one rotational DOF. Furthermore, criteria for evaluation of generated plans are introduced to reduce the candidates of manipulation plans. Fourth, a planning method for tying knots tightly is established because they fulfill their fixing function by tightening them. Finally, we report knotting/unknotting manipulation performed by a vision-guided system to demonstrate the usefulness of our approach.

Robust grasping manipulation of deformable objects

A simple but robust control law for the grasping manipulation of deformable objects is presented. In the handling of deformable objects, grasping and manipulation must be performed simultaneously despite uncertainties during the handling process. We propose a control law to perform grasping and manipulation of deformable objects using a real time vision system.

Rolling tensegrity driven by pneumatic soft actuators

In this paper, we describe the rolling of a tensegrity robot driven by a set of pneumatic soft actuators. Tensegrity is a mechanical structure consisting of a set of rigid elements connected by elastic tensional elements. Introducing tensegrity structures, we are able to build soft robots with larger size. Firstly, we show the prototype of a six-strut tensegrity robot, which is driven by twenty-four pneumatic McKibben actuators. Second, we formulate the geometry of the tensegrity robot. We categorize contact states between a six-strut tensegrity robot and a flat ground into two; axial symmetric contact and planar symmetric contact. Finally, we experimentally examine if rolling can be performed over a flat ground for individual sets of the actuators and discuss the strategy of rolling.

Crawling by body deformation of tensegrity structure robots

In this paper, we describe the design of a deformable robot with a tensegrity structure that can crawl and we show the results of experiments showing the ability of these robots to crawl. We first describe a tensegrity structure, composed of struts and cables, and its characteristics. We next explain the principle of crawling by robot body deformation, followed by a classification of the methods by which a body can be deformed and the contact conditions of the robot through the cable-graph of the tensegrity structure. We also describe topological transition graphs that can visualize crawling from each initial contact condition. We then discuss the characteristics of the proposed robot in terms of design freedom. Finally, we show experimentally that the prototype of a tensegrity robot can crawl.