Acquisition of a page turning skill for a multifingered hand using reinforcement learning

This paper proposes a method of acquisition of a page turning skill for a multifingered robotic hand using reinforcement learning. The goal of this paper is generation of the manipulation skill of a flexible object without its explicit model and tactile sensation during its control. In this paper, a page turning task is considered as an example of such tasks, where a sheet of paper, generally used in books, is a flexible object. The fingertip trajectories are obtained by reinforcement learning based on simulation. The reward considering a friction condition is given, so that a page turning skill without slip between the finger and the paper is obtained. The validity is confirmed by an experimental system which consists of a 2-d.o.f. manipulator and two 2-d.o.f. fingers.     

Biomimetic Pinching Movements of a Musculo-Skeletal Dual-Finger Model

The present paper investigates pinching movements using an index finger and a thumb actuated by redundant nonlinear digitorum muscles mimicking the configuration of human fingers. A dual-finger model with 2-d.o.f. joints for each finger and redundant nonlinear digitorum muscles is formulated to mimic the structure of human fingers. First, the kinematics and dynamics of the overall finger–object system, as well as the nonlinear muscular dynamics, are derived in accordance with the results of physiological studies. Next, a sensory-motor control law is proposed to enable stable pinching simultaneously with orientation regulation of an object. This control law includes an internal force term generated by co-construction of the redundant muscles. It is shown that the internal force term can modulate the damping factor in the joint space by its nonlinearity. Based on this effect, it is then shown by numerical simulation that our sensory-motor control law with co-contraction of each digitorum muscle makes it possible to realize pinching movements. Therefore, the pinching movements may be realized by means of a musculo-skeletal dual-finger system with the sensory-motor control law and co-contraction of redundant digitorum muscles.     

Keyswitch orientation can reduce finger joint torques during tapping on a computer keyswitch

OBJECTIVE: To examine the effects of keyswitch orientation on joint torques. BACKGROUND: The fingertip produces primarily vertical forces during single-finger tapping on a computer keyswitch. However, horizontal force components within the sagittal plane of the finger could reduce net joint torques. METHOD: Eleven participants tapped on a keyswitch oriented in three directions: vertical, tilted 30 degrees such that when pressed it moved away from the user (similar to a positive-tilt keyboard), and tilted 30 degrees such that when pressed it moved toward the user (similar to a negative-tilt keyboard). Participants also tapped on a prototype cantilever keyswitch design in which the key cap moves along the arc of a bending beam gradually away from the user. Miniature electro-optical goniometers measured the finger posture, and a two-axis force sensor measured fingertip forces. RESULTS: Tapping on a keyswitch oriented such that it moves away from the user when pressed reduced net joint torques by 47% relative to tapping on a vertically orientated keyswitch and by 56% relative to tapping on a keyswitch oriented toward the user, whereas the cantilever design resulted in 14% decreases in net joint torque relative to the vertical orientation. CONCLUSION: Reductions of torques resulted from decreasing the moment arm of the fingertip force about the joints. APPLICATION: Keyboard design should incorporate keyswitch mechanism angles along with other postural and geometric constraints to reduce exposure of the finger joints and muscles to force during typing.     

手指机构力操作灵巧性的分析与评价

本文对开端钢缆传动手指机构的力操作灵巧性问题进 行研究．首先,定义了钢缆张力椭球．在此基础上,分析了手指机构力操作的位形灵巧性及 钢缆传动结构对位形灵巧性的影响．其次,根据钢缆张力空间与指端操作力空间之间的映射 关系,给出了建立指端在给定操作位形处的操作力矢量集合的几何方法．得到的矢量集合反 映了考虑钢缆张力限制时,指端实际的力操作能力．为评价指端实际的力操作灵巧性,还提 出两个新的度量指标．最后,给出了手指机构力操作灵巧性分析的示例．     

A real-time strategy for dexterous manipulation: Fingertips motion planning,force sensing and grasp stability

This paper presents a global strategy for object manipulation with the fingertips with an anthropomorphic dexterous hand: the LMS Hand of the ROBIOSS team from PPRIME Institute in Poitiers (France). Fine manipulation with the fingertips requires to compute on one hand, finger motions able to produce the desired object motion and on the other hand, it is necessary to ensure object stability with a real time scheme for the fingertip force computation. In the literature, lot of works propose to solve the stability problem, but most of these works are grasp oriented; it means that the use of the proposed methods are not easy to implement for online computation while the grasped object is moving inside the hand. Also simple real time schemes and experimental results with full-actuated mechanical hands using three fingers were not proposed or are extremely rare. Thus we wish to propose in a same strategy, a robust and simple way to solve the fingertip path planning and the fingertip force computation. First, finger path planning is based on a geometric approach, and on a contact modelling between the grasped object and the finger. And as force sensing is required for force control, a new original approach based on neural networks and on the use of tendon-driven joints is also used to evaluate the normal force acting on the finger distal phalanx. And an efficient algorithm that computes fingertip forces involved is presented in the case of three dimensional object grasps. Based on previous works, those forces are computed by using a robust optimization scheme.In order to validate this strategy, different grasps and different manipulation tasks are presented and detailed with a simulation software, SMAR, developed by the PPRIME Institute. And finally experimental results with the real hand illustrate the efficiency of the whole approach.     

The design and control of a robot finger for tactile sensing

This article describes the design and control of a lightweight robot finger intended for tactile sensing research. The finger is a three-link planar chain with the joints actuated through cables by two motors. Kinematic coupling of the three joints provides two degrees of freedom for finger tip manipulation, and a curling action of the finger for enclosing an object. Hall effect sensors in each joint provide position feedback, and strain gage sensors on each cable provide tension information. To minimize weight and power consumption, a high speed low torque motor together with a 172:1 speed reducer is used as the actuator. A force control loop around the motor speed reducer system reduces the effect of the friction inherent in the speed reducer. Flat mounting plates are provided on each link for special purpose grasping surfaces and sensors.     

Ergonomic design of beverage can lift tabs based on numerical evaluations of fingertip discomfort

This paper introduces finite element analyses to evaluate numerically and objectively the feelings in the fingertip when opening aluminum beverage cans, in order to design the shape of the tab. Experiments of indenting vertically the fingertip pulp by a probe and by tabs of aluminum beverage can ends have allowed us to observe force responses and feelings in the fingertip. It was found that a typical force-displacement curve may be simplified as a combination of three curves with different gradients. Participants feel a touch at Curve 1 of the force-displacement curve, then feel a pressure and their pulse at Curve 2, finally feel discomfort followed by a pain in the fingertip at Curve 3. Finite element analyses have been performed to simulate indenting the tab with the fingertip vertically to confirm that the simulation results agree well with the experimental observations. Finally, numerical simulations of the finger pulling up the tab of the can end has also been performed and discomfort in the fingertip has been related to the maximum value of the contact stress of the finger model. Comparisons of three designs of tab ring shape showed that the tab with a larger contact area with finger is better.     

新型力解耦和各向同性机器人六维力传感器

介绍了基于并联 6 -SPS结构的新型机器人六维力与力矩传感器的结构特点 ,并分析计算了其力与应变之间的转换关系及其力各向同性 ,为该六维力与力矩传感器的设计和使用提供了理论依据。分析计算结果表明该传感器是力与力矩解耦和各向同性的。该传感器采用弹性铰链替代球铰 ,具有结构灵巧、工艺性好等优点 ,可应用到机器人手腕、手指和其它使用多维力与力矩传感器的场合。     

Effects of hand span size and right-left hand side on the piano playing performances: Exploration of the potential risk factors with regard to piano-related musculoskeletal disorders

This study used biomechanical techniques to objectively investigate the effects of the size of hand span and right/left hands on the kinematic and kinetic performances when playing the piano. Twenty pianists were recruited and assigned to two hand-size groups. The parameters of interest, such as ratio of maximal digit-to-digit abduction angle (RD-D_abd), range of motion (ROM) of finger and wrist flexion-extension (F/E), radial-ulnar deviation (R/U), and movement units of F/E of finger joints (MU_F/E), were measured while striking the piano keys. The fingertip force was also estimated by a kinetic model. The RD-D_abd was significantly larger for the small hand-span pianists when playing both chords and octaves. The ROM of wrist F/E was significantly larger for small hand-span pianists when playing chords. There was no significant difference in the fingertip force between two groups. However, the values for MU_F/E and fingertip force of the right hand were significantly larger than those of left hand. Pianists with a small hand-span should aware that they have higher exposure risks for hand injuries while playing the piano.Relevance to industry: Hand anthropometric issue might be one of potential risk factors which result in piano-related musculoskeletal disorders. This study provides preliminary evidence that can be used to aid in injury prevention and music education for pianists as well as to reconsider issues with regard to the piano design.     

一种基于空间连杆机构的蛇形机器人

提出了一种基于空间连杆机构的蛇形机器人模型．该蛇形机器人各骨节间通过万向铰链连接,每节安装两个微型伺服电机,电机带动曲柄转动,曲柄经两端为球铰的连杆驱动另一骨节运动．通过两个电机的耦合驱动,该蛇形机器人能实现蜿蜒爬行、转向、抬头等多种运动方式．文中给出了该空间连杆机构的运动学解,基于Serpenoid曲线计算得到了实际控制所需的关节转角和步数,并编程实现了上述运动方式．     

Kinematics for the whole arm of a serial-chain manipulator

This paper provides a viewpoint for kinematics for the whole arm of a serial-chain manipulator with 2-d.o.f. rotational joints. An in-depth understanding of the duality between a rigid link and a 2-d.o.f. joint allows us to derive simple and geometric equations describing the manipulator kinematics. The obtained kinematic equations are analyzed in two ways compared with the Frenet-Serret formula of a spatial curve which is utilized for a reference shape of the manipulator. One way is based on limit analysis where we increase the number of joints while the total length of the manipulator remains constant. The other way utilizes an extended mechanism through the link-joint duality. The information presented in this paper is useful for mechanism design dynamic analysis, control design and motion planning of the manipulators in whole-arm manipulation.     

SMA丝驱动的手指康复机器人设计及实验研究

为了研制一种轻便、可穿戴的仿生手指康复机器人,在分析手指肌肉骨骼生物参数及致动机理的基础上,设计了一种形状记忆合金丝(Shape Memory Alloy, SMA)驱动的软体柔性手指康复机器人,建立了其运动学和力学模型。以手套为结构设计的原型,通过控制SMA丝的收缩来模拟手指肌肉肌腱的收缩,从而实现辅助手指屈曲伸展运动的功能。试验研究了手指康复机器人的运动性能和抓握性能。试验结果表明,手指柔性康复机器人最大弯曲角度接近正常人手关节角度,最大指尖力可达18N,能完成日常所需的屈曲伸展以及抓握功能。     

Antagonistic muscle-like actuator and its application to multi-d.o.f. forearm prosthesis

It is first clarified through simulational analysis that co-activation between agonist an muscle and its antagonist both having nonlinear elasticity is indispensable for the stiffness adjustment on an articulator movement. An antagonistic muscle-like actuator (AMA) is then proposed as an approach of imitating a skeleto-muscular articulation system. The AMA has two nonlinear springs which drive one joint antagonistically and provides the joint with the feasibility of stiffness adjustment. The development of the forearm prosthesis which has 6 d.o.f. is developed as an application of the AMA. Two d.o.f. for the wrist and 4 d.o.f. for the two fingers are driven by only one DC motor and one stepping motor for exchanging the driving mode. The whole mechanism is designed as almost the same volume as an adult's forearm. A shape fit mechanism for the finger is proposed, by which the three joints of one finger automatically adjust their angle according to the shape of the object to be grasped.     

一种单电机驱动多指多关节机械手的设计

为了使机械手灵巧、稳妥地抓取物体,设计了一种新型结构的单电机驱动4 指12 关节机械手爪．该手 爪由电机驱动一根十字连杆,其端部分别连接4 个手指的第1 动力连杆;每个手指有3 个指节,由2 个平行四边形 的指节结构确保手指末端做平移运动;每个手指的第2 动力连杆具有延伸滑槽,当第2 动力连杆运动时,经过特别 设计的滑槽在固定支点滑动,可使手指末端匀速运动．该新型的单电机驱动手爪设计方案实现了机械手控制简单、 抓握可靠的目的．     

Pneumatic impedance control of a 3-d.o.f. physiotherapy robot

Stroke is a common condition resulting in 30 000 people per annum left with significant disability. In patients with severe arm paresis after stroke, functional recovery in the affected arm is poor. Inadequate intensity of treatment is cited as one factor accounting for the lack arm recovery found in some studies. Given that physical therapy resource is limited, strategies to enhance the physiotherapists' efforts are needed. One approach is to use robotic techniques to augment movement therapy. A 3-d.o.f. pneumatic robot has been developed to apply physiotherapy to the upper limb. The robot has been designed with a workspace encompassing the reach-retrieve range of the average male. Slight non-linearities in the response of the pneumatic system are observed and explained. Building upon previous work that used an error-prone custom force sensor, a commercial 6-d.o.f. force sensor is used to apply impedance control in 3 d.o.f. on the robot.     

Management of parallel-manipulator singularities using joint-coupling

Joint-coupling (JC) is introduced in the design and control of parallel manipulators for managing manipulator singularities. The underlying idea consists of coupling the motions of several manipulator joints by driving the joints with a single actuator. This can be done by mechanical or electronic means, while preserving the mobility of the manipulator. Such JC is intended to alter the direct singularity condition of parallel manipulators upon changing the coupling coefficients. Both 2- and 3-d.o.f. planar parallel manipulator examples are provided to illustrate the concept of JC and singularity management. The prototype of a JC device and its associated 2-d.o.f. parallel manipulator intended for a feasibility study is introduced.     

Stabilization control for biped follow walking

This paper proposes a follow-walking motion for biped humanoid robots based on a stabilization control and a complete walking-motion pattern. To follow human motion, the unit patterns of the trunk, the waist and the lower limbs are generated and synthesized. During the follow motion, the biped robots are balanced by the stabilization control that calculates the combined motion of the trunk and the waist that compensates for the moments produced by the motion of the lower limbs. For confirmation of the follow-walking motion, we have developed a life-sized humanoid robot, WABIAN-RII (WAseda BIped humANoid robot-Revised II). It has a total of 43 mechanical d.o.f.; two 6-d.o.f. legs, two 10-d.o.f. arms, a 4-d.o.f. neck, 4-d.o.f. in the eyes and a torso with a 3-d.o.f. waist.     

Friction model of fingertip sliding over wavy surface for friction-variable tactile feedback panel

A friction-variable touch panel is capable of presenting virtual bumps and holes on its flat surface through the control of the surface friction when a fingertip slides over it. To improve the presentation, we developed a friction model of a fingertip sliding over a sinusoidal surface with an amplitude of 0.5–2.5 mm and a spatial wavelength of 20–50 mm. When a metal ball rolls over a wavy surface with a low friction and contact area, the ratio of the horizontal force to the normal force is equal to the gradient of the surface (this is referred to as the ball bearing model) and is hardly affected by the normal load and rolling speed. In contrast, the profile of the force ratio of a sliding finger is substantially skewed and affected by the sliding direction and normal force exerted by the finger. To model this skewed force ratio, we formulated the asymmetric pressure distribution in the finger-surface contact area and used the effects of the adhesion friction to model the dependency of the force ratio on the normal force and sliding direction. The developed model of a bare finger with these features was found to sufficiently simulate the experimentally observed force ratios. The model can be easily applied to friction-variable touch panels and enables the achievement of a wide variety of haptic contents with macroscopically concave or convex surfaces.     

Two-finger stiffness discrimination of a virtual object with haptic sensation enhancement via the stochastic resonance effect

We investigated the stiffness discrimination ability of two fingers with stochastic resonance. Haptic perception at the fingertip is known to improve when vibrotactile noise propagates to the fingertip, and this phenomenon is called stochastic resonance. The improvement of a fingertip’s haptic sensation depends on the intensity of the noise propagating to the fingertip. Therefore, to improve the haptic sensation at multiple fingertips, we do not need to attach noise sources, such as vibrators, to multiple fingertips. Even if only one vibrator is used, the haptic sensation at multiple fingertips may be improved by propagating noise to the fingertips. In this study, we focus on stiffness discrimination as a task using multiple fingers, in which the thumb and index finger are used to touch a virtual object and perceive its stiffness. Subsequently, we demonstrate that the stiffness perception is improved by propagating sufficiently intense noise to the thumb and index finger using a vibrator. Furthermore, this study considers the same task in a real-world situation to investigate the relationship between the stochastic resonance effect and the stiffness discrimination task. The findings indicate the possibility of improving the haptic sensation at multiple fingertips using one vibrator.     

Design of an optical soft sensor for measuring fingertip force and contact recognition

Among the various ways to estimate user intention in hand exoskeletons, a contact force measurement is definitely the most straightforward and intuitive method. A force sensor, located at the center of a fingertip usually, hinders the tactile sensation of the user by blocking the contact between an object and the fingertip. To overcome this problem, a soft force sensor with horse shoe shape is utilized to measure the contact force and provide the tactile sensation to the user. This work presents the mechanical design, implementation and evaluation of a soft fingertip force sensor. To maximize tactile sensation of the user, we adopted a horse shoe shape structure to leave the finger pad exposed. An optical sensing mechanism was selected for its relatively fast response compared to other soft sensors. The whole sensor system has a soft exterior providing flexibility and a user-friendly interface. To evaluate the sensor’s performance, we carried out sensor optimization process and calibration experiment with a customized test bed. Then, we investigated both static and dynamic response and observed the mechanical behavior and light intensity changes caused by the cross sectional shape and base/agent ratio of PDMS. Lastly, we applied the proposed sensor to the glove type fingertip force monitoring system. The sensor estimates the index finger tip force with high accuracy (R ² = 0:96) within 5N range.