Anthropomorphic Robotic Hand Research Articles

Stretchable glove for accurate and robust hand pose reconstruction based on comprehensive motion data

We propose a compact wearable glove capable of estimating both the finger bone lengths and the joint angles of the wearer with a simple stretch-based sensing mechanism. The soft sensing glove is designed to easily stretch and to be one-size-fits-all, both measuring the size of the hand and estimating the finger joint motions of the thumb, index, and middle fingers. The system was calibrated and evaluated using comprehensive hand motion data that reflect the extensive range of natural human hand motions and various anatomical structures. The data were collected with a custom motion-capture setup and transformed into the joint angles through our post-processing method. The glove system is capable of reconstructing arbitrary and even unconventional hand poses with accuracy and robustness, confirmed by evaluations on the estimation of bone lengths (mean error: 2.1 mm), joint angles (mean error: 4.16°), and fingertip positions (mean 3D error: 4.02 mm), and on overall hand pose reconstructions in various applications. The proposed glove allows us to take advantage of the dexterity of the human hand with potential applications, including but not limited to teleoperation of anthropomorphic robot hands or surgical robots, virtual and augmented reality, and collection of human motion data.

Anthropomorphic Robotic Hand Prosthesis Developed for Children.

The use of both hands is a common practice in everyday life. The capacity to interact with the environment is largely dependent on the ability to use both hands. A thorough review of the current state of the art reveals that commercially available prosthetic hands designed for children are very different in functionality from those developed for adults, primarily due to prosthetic hands for adults featuring a greater number of actuated joints. Many times, patients stop using their prosthetic device because they feel that it does not fit well in terms of shape and size. With the idea of solving these problems, the design of HandBot-Kid has been developed with the anthropomorphic qualities of a child between the ages of eight and twelve in mind. Fitting the features of this age range, the robotic hand has a length of 16 cm, width of 7 cm, thickness of 3.6 cm, and weight of 328 g. The prosthesis is equipped with a total of fifteen degrees of freedom (DOF), with three DOFs allocated to each finger. The concept of design for manufacturing and assembly (DFMA) has been integrated into the development process, enabling the number of parts to be optimized in order to reduce the production time and cost. The utilization of 3D printing technology in conjunction with aluminum machining enabled the manufacturing process of the robotic hand prototype to be streamlined. The flexion-extension movement of each finger exhibits a trajectory that is highly similar to that of a real human finger. The four-bar mechanism integrated into the finger design achieves a mechanical advantage (MA) of 40.33% and a fingertip pressure force of 10.23 N. Finally, HandBot-Kid was subjected to a series of studies and taxonomical tests, including Cutkosky (16 points) and Kapandji (4 points) score tests, and the functional results were compared with some commercial solutions for children mentioned in the state of the art.

Learning-Based Planner for Unknown Object Dexterous Manipulation Using ANFIS

Dexterous manipulation of unknown objects performed by robots equipped with mechanical hands represents a critical challenge. The difficulties arise from the absence of a precise model of the manipulated objects, unpredictable environments, and limited sensing capabilities of the mechanical hands compared to human hands. This paper introduces a data-driven approach that provides a learning-based planner for dexterous manipulation employing an Adaptive Neuro-Fuzzy Inference System (ANFIS) fed by data obtained from an analytical manipulation planner. ANFIS captures the complex relationships between inputs and optimal manipulation parameters. Moreover, during a training phase, it is able to fine-tune itself on the basis of its experiences. The proposed planner enables a robot to interact with objects of various shapes, sizes, and material properties while providing an adaptive solution for increasing robotic dexterity. The planner is validated in a real-world environment, applying an Allegro anthropomorphic robotic hand. A link to a video of the experiment is provided in the paper.

Towards humanlike grasp in robotic hands: mechanical implementation of force synergies

In the field of robotic hands, finger force coordination is usually achieved by complex mechanical structures and control systems. This study presents the design of a novel transmission system inspired from the physiological concept of force synergies, aiming to simplify the control of multifingered robotic hands. To this end, we collected human finger force data during six isometric grasping tasks, and force synergies (i.e. the synergy weightings and the corresponding activation coefficients) were extracted from the concatenated force data to explore their potential for force modulation. We then implemented two force synergies with a cable-driven transmission mechanism consisting of two spring-loaded sliders and five V-shaped bars. Specifically, we used fixed synergy weightings to determine the stiffness of the compression springs, and the displacements of sliders were determined by time-varying activation coefficients. The derived transmission system was then used to drive a five-finger robotic hand named SYN hand. We also designed a motion encoder to selectively activate desired fingers, making it possible for two motors to empower a variety of hand postures. Experiments on the prototype demonstrate successful grasp of a wide range of objects in everyday life, and the finger force distribution of SYN hand can approximate that of human hand during six typical tasks. To our best knowledge, this study shows the first attempt to mechanically implement force synergies for finger force modulation in a robotic hand. In comparison to state-of-the-art robotic hands with similar functionality, the proposed hand can distribute humanlike force ratios on the fingers by simple position control, rather than resorting to additional force sensors or complex control strategies. The outcome of this study may provide alternatives for the design of novel anthropomorphic robotic hands, and thus show application prospects in the field of hand prostheses and exoskeletons.

Concept and Design of Anthropomorphic Robot Hand with a Finger Movement Mechanism based on a Lever for Humanoid Robot T-FLoW 3.0

This work described a concept and design of an anthropomorphic robot hand for the T-FLoW 3.0 humanoid robot, which featured a mechanism based on a lever as its finger movement. This work aimed to provide an affordable, modular, lightweight, human-like robot hand with a mechanism that minimizes mechanical slippage. The proposed mechanism works based on the push/pull of a lever attached to the finger to generate its finger flexion/extension movement. The finger’s lever is pushed/pulled through a servo horn and a rigid bar by the affordable TowerPro MG90S micro-servo. Our hand is developed only as necessary to become close to human hands by only applying five fingers and six joints, where each joint has its actuator. The combination of 3D printing technology with PLA filament accelerates and streamlines the manufacturing process, provides a realistic appearance, and achieves a lightweight, affordable, and easy maintenance product. Structural analysis simulations show that our finger design constructed with PLA material could withstand a load of about 30 N. We verified our finger mechanism by repeatedly flexing and extending the finger 30 times, and the results showed that the finger movements could be performed well. Our hand offered excellent handling for the mechanical issues brought on by finger movements, one of the issues that robot hand researchers have encountered. Our work could provide significant benefits to the T-FLoW 3.0 developers in enhancing the ability of humanoid robots involving hands, such as grasping and manipulating objects.

A Self-sensing TSA-actuated Anthropomorphic Robot Hand

A Self-sensing TSA-actuated Anthropomorphic Robot Hand

Teleoperation of an Anthropomorphic Robot Hand with a Metamorphic Palm and Tunable-Stiffness Soft Fingers.

Teleoperation in soft robotics can endow soft robots with the ability to perform complex tasks through human-robot interaction. In this study, we propose a teleoperated anthropomorphic soft robot hand with variable degrees of freedom (DOFs) and a metamorphic palm. The soft robot hand consists of four pneumatic-actuated fingers, which can be heated to tune stiffness. A metamorphic mechanism was actuated to morph the hand palm by servo motors. The human fingers' DOF, gesture, and muscle stiffness were collected and mapped to the soft robotic hand through the sensory feedback from surface electromyography devices on the jib. The results show that the proposed soft robot hand can generate a variety of anthropomorphic configurations and can be remotely controlled to perform complex tasks such as primitively operating the cell phone and placing the building blocks. We also show that the soft hand can grasp a target through the slit by varying the DOFs and stiffness in a trail.

A marker-based approach to determine the centers of rotation of finger joints

Background and ObjectiveThe methods proposed in literature to estimate the position of hand joints Centers of Rotation (CoRs) typically require computationally non-trivial optimization routines and exploit a high number of markers to calculate CoRs positions from surface marker trajectories. Moreover, most of the existing works evaluated the accuracy only in simulation. This work proposes a new procedure, based on the Pratt circle fit, to estimate joints CoRs position in 2D through marker-based acquisitions. MethodsThe advantage of the Pratt circle fit lies in its simplicity and computational speed, and in the possibility of exploiting a reduced markerset for calculating CoRs. By applying simplifying assumptions regarding the movement of the fingers (i.e., planar and decoupled flexion-extension movements of each joint occurring in the same flexion plane for all the joints of the finger), it is possible to determine the position of the CoR of each joint in 2D. For this reason, the estimation of the Carpo-MetaCarpal joint of the thumb was not included in this work, as it exhibits a more complex movement associated to the combination of a flexion-extension and adduction-abduction degree of freedom. The errors in estimating CoRs were evaluated by conducting experimental acquisitions on an anthropomorphic robotic hand and comparing the position of the estimated CoR with the real position of the CoR. The repeatability of the method and its capability to estimate anatomically plausible CoRs were evaluated through experimental acquisitions conducted on five healthy volunteers. ResultsErrors in estimating finger joints CoRs were in the order of 0.70 mm and 0.18 mm respectively along the finger longitudinal direction (i.e., x coordinate) and thickness (i.e., y coordinate). Standard Deviations of CoRs positions were comparable to the ones obtained in literature (i.e., below 2 mm and 1 mm respectively for the x and y coordinates), thus demonstrating the repeatability of the method. The Anatomical Plausibility Rate of the proposed approach was between 80% and 100%. ConclusionsThe performance of the Pratt-based CoRs estimation procedure proposed in this work was comparable to other existing methods, with the advantage of exploiting a simple fitting algorithm and a reduced markerset with respect to the state-of-the-art techniques.

Deep Learning-Based Ensemble Approach for Autonomous Object Manipulation with an Anthropomorphic Soft Robot Hand

Autonomous object manipulation is a challenging task in robotics because it requires an essential understanding of the object’s parameters such as position, 3D shape, grasping (i.e., touching) areas, and orientation. This work presents an autonomous object manipulation system using an anthropomorphic soft robot hand with deep learning (DL) vision intelligence for object detection, 3D shape reconstruction, and object grasping area generation. Object detection is performed using Faster-RCNN and an RGB-D sensor to produce a partial depth view of the objects randomly located in the working space. Three-dimensional object shape reconstruction is performed using U-Net based on 3D convolutions with bottle-neck layers and skip connections generating a complete 3D shape of the object from the sensed single-depth view. Then, the grasping position and orientation are computed based on the reconstructed 3D object information (e.g., object shape and size) using U-Net based on 3D convolutions and Principal Component Analysis (PCA), respectively. The proposed autonomous object manipulation system is evaluated by grasping and relocating twelve objects not included in the training database, achieving an average of 95% successful object grasping and 93% object relocations.

Myo Transformer Signal Classification for an Anthropomorphic Robotic Hand

The evolution of anthropomorphic robotic hands (ARH) in recent years has been sizable, employing control techniques based on machine learning classifiers for myoelectric signal processing. This work introduces an innovative multi-channel bio-signal transformer (MuCBiT) for surface electromyography (EMG) signal recognition and classification. The proposed MuCBiT is an artificial neural network based on fully connected layers and transformer architecture. The MuCBiT recognizes and classifies EMG signals sensed from electrodes patched over the arm’s surface. The MuCBiT classifier was trained and validated using a collected dataset of four hand gestures across ten users. Despite the smaller size of the dataset, the MuCBiT achieved a prediction accuracy of 86.25%, outperforming traditional machine learning models and other transformer-based classifiers for EMG signal classification. This integrative transformer-based gesture recognition promises notable advancements for ARH development, underscoring prospective improvements in prosthetics and human–robot interaction.

Fully 3D printed flexible, conformal and multi-directional tactile sensor with integrated biomimetic and auxetic structure

Tactile sensors play a crucial role in the development of biologically inspired robotic prostheses, particularly in providing tactile feedback. However, existing sensing technology still falls short in terms of sensitivity under high pressure and adaptability to uneven working surfaces. Furthermore, the fabrication of tactile sensors often requires complex and expensive manufacturing processes, limiting their widespread application. Here we develop a conformal tactile sensor with improved sensing performance fabricated using an in-house 3D printing system. Our sensor detects shear stimuli through the integration of an auxetic structure and interlocking features. The design enables an extended sensing range (from 0.1 to 0.26 MPa) and provides sensitivity in both normal and shear directions, with values of 0.63 KPa−1 and 0.92 N−1, respectively. Additionally, the sensor is capable of detecting temperature variations within the range of 40−90 °C. To showcase the feasibility of our approach, we have printed the tactile sensor directly onto the fingertip of an anthropomorphic robotic hand, the proximal femur head, and lumbar vertebra. The results demonstrate the potential for achieving sensorimotor control and temperature sensing in artificial upper limbs, and allowing the monitoring of bone-on-bone load.

Dexterous all-soft hand (DASH) with active palm: multi-functional soft hand beyond grasping

Safety and adaptability are key features of anthropomorphic robotic hands and tend to receive significant attention, however, the motion of the palm, which is also an essential element of a dexterous hand, is often ignored. Not only does a rigid palm lack compliance, but its limited manipulative capacity restricts the soft anthropomorphic hand’s range of applications. Thus, in this study, we present the dexterous all-soft hand (DASH), featuring an active palm based on pneumatic actuation. The DASH is comprised of five fingers and an active palm with 11 degrees of freedom, which enable it to achieve a range of metacarpal motions as well as thumb opposition. A step-by-step casting method was used to fabricate the hand and a set of tests were performed to characterize the bending angle and the blocked force between fingers and palm. The ability of this novel hand to achieve the highest possible scores in the Kapandji test, the injection operation, scissors manipulation and unscrewing the cap was subsequently demonstrated. The DASH is the same size as an average adult female hand, and offers the additional benefits of light weight (165 g), low cost and ease of fabrication. Moreover, its manipulation ability further broadens the range of potential applications for this type of robotic hand.

Ion trap and release dynamics enables nonintrusive tactile augmentation in monolithic sensory neuron.

An iontronic-based artificial tactile nerve is a promising technology for emulating the tactile recognition and learning of human skin with low power consumption. However, its weak tactile memory and complex integration structure remain challenging. We present an ion trap and release dynamics (iTRD)-driven, neuro-inspired monolithic artificial tactile neuron (NeuroMAT) that can achieve tactile perception and memory consolidation in a single device. Through the tactile-driven release of ions initially trapped within iTRD-iongel, NeuroMAT only generates nonintrusive synaptic memory signals when mechanical stress is applied under voltage stimulation. The induced tactile memory is augmented by auxiliary voltage pulses independent of tactile sensing signals. We integrate NeuroMAT with an anthropomorphic robotic hand system to imitate memory-based human motion; the robust tactile memory of NeuroMAT enables the hand to consistently perform reliable gripping motion.

Dexterous manipulation of construction tools using anthropomorphic robotic hand

Emerging studies are utilizing reinforcement learning (RL) and imitation learning (IL) to control large-scale robots in heavy construction tasks. There is limited attention given to the automation of delicate tasks typically performed manually. This paper proposes the control of an anthropomorphic robotic hand with a high degree of freedom for the manipulation of construction tools in a learning-based approach. For controlling the robotic hand, a simulation-based policy learning framework based on pretraining policies through IL is presented, subsequently fine-tuning them for construction tool manipulation using RL. In experimental trials, six policies are trained for the robotic hand to grasp six different construction tools. The results indicate that each of the learned policies enables the robotic hand to manipulate the corresponding tool with an almost 100% success rate, demonstrating resilience when confronted with tools of different scales. Additionally, the paper showcases the potential for scaling up the fundamental policy for downstream applications.

Learning Human-to-Robot Dexterous Handovers for Anthropomorphic Hand

Human-robot interaction plays an important role in robots serving human production and life. Object handover between humans and robotics is one of the fundamental problems of human-robot interaction. The majority of current work uses parallel-jaw grippers as the end-effector device, which limits the ability of the robot to grab miscellaneous objects from human and manipulate them subsequently. In this paper, we present a framework for human-to-robot dexterous handover using an anthropomorphic hand. The framework takes images captured by two cameras to complete handover scene understanding, grasp configurations prediction, and handover execution. To enable the robot to generalize to diverse delivered objects with miscellaneous shapes and sizes, we propose an anthropomorphic hand grasp network (AHG-Net), an end-to-end network that takes the single-view point clouds of the object as input and predicts the suitable anthropomorphic hand configurations with 5 different grasp taxonomies. To train our model, we build a large-scale dataset with 1M hand grasp annotations from 5K single-view point clouds of 200 objects. We implement a handover system using a UR5 robot arm and HIT-DLR II anthropomorphic robot hand based on our presented framework, which can not only adapt to different human givers but generalize to diverse novel objects with various shapes and sizes. The generalizability, reliability, and robustness of our method are demonstrated on 15 different novel objects with arbitrary handover poses from frontal and lateral positions, a system ablation study, a grasp planner comparison, and a user study on 6 participants delivering 15 objects from two benchmark sets.

An Affordable Telerobotic System Architecture for Grasp Training and Object Grasping for Human–Machine Interaction

Abstract Due to mobility impairment, a person might rely on wheelchairs, canes, and crutches for assistance but could face challenges when performing tasks such as grasping and manipulating objects due to limitations in reach and capability. To overcome these challenges, a multidegree-of-freedom robotic arm with an anthropomorphic robotic hand (ARH) could be used. In this research, we propose an architecture and then implement it toward the development of an assistive system to assist a person with object grasping. The architecture interlinks three functional modules to provide three operation modes to calibrate the system, train a user on how to execute a grasp, synthesize grasps, and execute a grasp. The developed system consists of a user input and feedback glove capable of capturing user inputs and providing grasp-related vibrotactile feedback, a coppeliasim virtual environment emulating the motions of the ARH, and an underactuated ARH capable of executing grasps while sensing grasp contact locations. The operation of the developed system is evaluated to determine the ability of a person to operate it and perform a grasp using two control methods; using a synthesized grasp or under real-time continuous control. The successful evaluation validates the architecture and the developed system to provide the ability to perform a grasp. The results of the evaluation provide confidence in expanding the system capabilities and using it to develop a database of grasp trajectories of objects with different geometries.

An Intelligent Robotic System Capable of Sensing and Describing Objects Based on Bimodal, Self‐Powered Flexible Sensors

AbstractThis study presents an intelligent soft robotic system capable of perceiving, describing, and sorting objects based on their physical properties. This work introduces a bimodal self‐powered flexible sensor (BSFS) based on the triboelectric nanogenerator and giant magnetoelastic effect. The BSFS features a simplified structure comprising a magnetoelastic conductive film and a packaged liquid metal coil. The BSFS can precisely detect and distinguish touchless and tactile models, with a response time of 10 ms. By seamlessly integrating the BSFSs into the soft fingers, this study realizes an anthropomorphic soft robotic hand with remarkable multimodal perception capabilities. The touchless signals provide valuable insights into object shape and material composition, while the tactile signals offer precise information regarding surface roughness. Utilizing a convolutional neural network (CNN), this study integrates all sensing information, resulting in an intelligent soft robotic system that accurately describes objects based on their physical properties, including materials, surface roughness, and shapes, with an accuracy rate of up to 97%. This study may lay a robotic foundation for the hardware of the general artificial intelligence with capacities to interpret and interact with the physical world, which also serves as an interface between artificial intelligence and soft robots.

Dream to posture: visual posturing of a tendon-driven hand using world model and muscle synergies

Although tendon-driven anthropomorphic robot hands have the potential to achieve human-level dexterity, controlling them is a great challenge owing to their mechanical complexities. Therefore, investigating human-hand control strategies is of the utmost importance. An important skill that enables the versatile manipulation ability of humans is visual posturing, i.e. the skill to make arbitrary hand postures based solely on visual observation. Visual posturing facilitates manipulation learning by enabling visual imitation learning and reusing visually similar past experiences. Therefore, this study investigates a method to replicate visual posturing in anthropomorphic robotic hands. Visual posturing in tendon-driven hands is challenging because of the hysteresis in tendon systems, the partial observability of the problem, and the presence of many actuators owing to the antagonistic tendon arrangement. To address these challenges, we propose a method that combines a model predictive path integral, a world model, and bio-inspired muscle synergies. The evaluation in a physical tendon-driven anthropomorphic robot hand showed that the proposed method achieved better visual posturing performance than a naive regression model. We anticipate that our visual posturing method will lay the foundation for versatile manipulation controllers that can adaptively learn manipulation tasks, similar to humans.

ROS End-Effector: A Hardware-Agnostic Software and Control Framework for Robotic End-Effectors

In recent years, several robotic end-effectors have been developed and made available in the market. Nevertheless, their adoption in industrial context is still limited due to a burdensome integration, which strongly relies on customized software modules specific for each end-effector. Indeed, to enable the functionalities of these end-effectors, dedicated interfaces must be developed to consider the different end-effector characteristics, like finger kinematics, actuation systems, and communication protocols. To face the challenges described above, we present ROS End-Effector, an open-source framework capable of accommodating a wide range of robotic end-effectors of different grasping capabilities (grasping, pinching, or independent finger dexterity) and hardware characteristics. The ROS End-Effector framework, rather than controlling each end-effector in a different and customized way, allows to mask the physical hardware differences and permits to control the end-effector using a set of high-level grasping primitives automatically extracted. By leveraging on hardware agnostic software modules including hardware abstraction layer (HAL), application programming interfaces (APIs), simulation tools and graphical user interfaces (GUIs), ROS End-Effector effectively facilitates the integration of diverse end-effector devices. The proposed framework capabilities in supporting different robotics end-effectors are demonstrated in both simulated and real hardware experiments using a variety of end-effectors with diverse characteristics, ranging from under-actuated grippers to anthropomorphic robotic hands. Finally, from the user perspective, the manuscript provides a set of examples about the use of the framework showing its flexibility in integrating a new end-effector module.

An Anthropomorphic Robotic Hand With a Soft-Rigid Hybrid Structure and Positive- Negative Pneumatic Actuation

Anthropomorphic robotic hands are seeking to achieve key features such as multi-degree-of-freedom motion ability, bi-directional actuation, high adaptability, and sufficient stiffness. In this research, we propose a 10 active degrees-of-freedom anthropomorphic robotic hand with a soft-rigid hybrid structure and positive-negative pneumatic actuation. The fingers of the hand utilize pneumatic actuators composed of soft bellows to actuate a rigid skeleton for bending motion with unidirectional-stretchable nylon elastic fabric as a strain limiting layer. The positive-negative pneumatic actuation mimics the flexors and extensors of the human hand to implement bi-directional finger actuation. We present a prototype implementation of this robotic hand and provide a preliminary evaluation. The experimental results show that the combined positive-negative pneumatic actuation can extend the workspace of the finger and increase the finger stiffness compared to using only positive pneumatic actuation. The proposed robotic hand has good comprehensive performance (generated a blocking forceof 7.5 N, scored 7 in theKapandji test, achieved32 of the 33 grasp postures from the Feix taxonomy, and could implement 5 human grasping strategies) compared to other robotic hands in literatures.