Recent Progress in Advanced Tactile Sensing Technologies for Soft Grippers

Tactile sensing technology is crucial for soft grippers. Soft grippers equipped with intelligent tactile sensing systems based on various sensors can interact safely with the unstructured environments and obtain precise properties of objects (e.g., size and shape). It is essential to develop state-of-the-art sensing technologies for soft grippers to handle different grasping tasks. In this review, the development of tactile sensing techniques for robotic hands is first introduced. Then, the principles and structures of different types of sensors normally adopted in soft grippers, including capacitive tactile sensors, piezoresistive tactile sensors, piezoelectric tactile sensors, fiber Bragg grating (FBG) sensors, vision-based tactile sensors, triboelectric tactile sensors, and other advanced sensors developed recently are briefly presented. Furthermore, sensing modalities and methodologies for soft grippers are also described in aspects of force measurement, perception of object properties, slip detection, and fusion of perception. The application scenarios of soft grippers are also summarized based on these advanced sensing technologies. Finally, the challenges of tactile sensing technologies for soft grippers that need to be tackled are discussed and perspectives in addressing these challenges are pointed out.     

基于压电薄膜的多信息触觉检测系统

为模拟人类触觉感知功能,实现智能触觉,研制了一种基于聚偏氟乙烯（PVDF）压电薄膜的多信息触觉检测系统,初步实现了对物体表面柔软度、黏性、纹理粗糙度、纹理规律性、热度5种维度信息的检测识别。该触觉检测系统利用凸点型压电薄膜传感器获取物体表面信息,上位机对采样信号进行数字滤波、特征提取及特征分类,最终得到物体表面特性信息,并绘出物体表面特性五感图。其中纹理粗糙度采用主频率识别方法,利用傅里叶变换（FFT）进行特征提取。实验结果表明,该系统能够有效区分被测物体表面信息。     

An intelligent robotic system based on a fuzzy approach

This paper deals with a fuzzy-based intelligent robotic system that requires various capabilities normally associated with intelligence. It acquires skills and knowledge through interaction with a dynamic environment. Subsumption architectures, behavior-based artificial intelligence, and behavioral engineering for robotic systems have been discussed as new technologies for intelligent robotic systems. This paper proposes a robotic system with “structured intelligence”. We focus on a mobile robotic system with a fuzzy controller and propose a sensory network that allows the robot to perceive its environment. An evolutionary approach improves the robot's performance. Furthermore, we discuss the effectiveness of the proposed method through computer simulations of collision avoidance and path-planning problems     

Stretchable Skin‐Like Cooling/Heating Device for Reconstruction of Artificial Thermal Sensation in Virtual Reality

Along with visual and tactile sensations, thermal sensation by temperature feeling on the skin can provide rich physical information on the environment and objects. With a simple touch of objects, relative temperature can be sensed and even objects can be differentiated with different thermal properties without any visual cue. Thus, artificially reproducing accurate/controllable thermal sensation haptic signals on human epidermis will certainly be a major research area to reconstruct a more realistic virtual reality (VR) environment. In this study, for the first time, a skin‐like, highly soft and stretchable and bi‐functional (both cold and hot sensation) thermo‐haptic device is reported for wearable VR applications with a single device structure (not separate heater and cooler). The skin‐like thermo‐haptic (STH) device can actively cool down and heat up deformable skin surfaces with instantaneous and accurate adjustment of temperature based upon a feedback control algorithm to mimic desirable thermal sensation with 230% stretchability. As a proof‐of‐concept, the STH device is integrated with a finger‐motion tracking glove to provide artificial thermal sensation information to the skin in various situations such as touching cold beer bottles and hot coffee cups in virtual space. This new type of STH device can offer potential implications for next‐generation haptic devices to provide unique thermal information for a more realistic virtual‐world field and medical thermal treatment.     

Progresses in Tensile,Torsional, and Multifunctional Soft Actuators

Soft actuators have received intensive attention in the fields of soft robots, sensors, intelligent control, artificial intelligence, and visual intelligence. By combination of tensile and torsional deformations, different types of motions can be realized, such as bending, rolling, and jumping. Soft robotics need soft actuators, such as artificial muscles to lift or move objects to perform some work. Additionally, actuation integrated with functions of sensing, signal transmission, and control is also needed in the development of advanced intelligent systems, which further stimulates the requirement of multifunctional actuators. Here different types of soft actuators that can perform tensile and torsional actuations are summarized, including twisted fiber artificial muscles, shape memory polymers, hydrogels, liquid crystal polymers, electrochemical actuators with conducting polymers, and some natural materials. Examples are also included regarding the bending or rolling deformations of the actuators for lifting objects. Then, recent interesting reports about multifunctional soft actuators combined with sensing and signal transmission performances, are summarized. Last, a summary of different ways to realize tensile and torsional actuations, different materials, and designs for lifting or moving objects, as well as construction of multifunctional actuators with actuation and sensing functions is provided.     

智能电磁感知的若干进展

智能电磁感知是电磁探测与成像的系统化和智能化延伸,是安全检查、生物医学、物联网等领域的基础性、关键性和共性问题.近年来,挖掘利用人工电磁材料和人工智能在电磁波调控与数据信息调控方面的强大能力,将其有机结合,并系统地引入电磁感知领域,发展了低成本、高性能的智能电磁感知体制,为电磁感知的进一步发展提供了关键理论和技术支撑....     

Telesurgery Robot Based on 5G Tactile Internet

With the development of modern medical technology, the emerging 5G, tactile Internet, robot, and artificial intelligence technology have enabled the interdisciplinary innovations facilitating the development of the surgical treatment technology, and enhancing the treatment efficiency of various diseases. In the medical field, the introduction of robot technology has contributed to the telesurgery. Moreover, the telesurgery robot allocated with the 5G tactile Internet as infrastructure, and AI technology as core competitiveness can promote the audio, visual and tactile perceptions of a doctor during the surgery process and solve the problems of resource scheduling; accordingly, it has become the research hotspot. Therefore, this paper introduces a telesurgery robot based on the 5G tactile Internet and artificial intelligence technology. The architecture, composition, characteristics, and advantages of telesurgery are explained in detail from two aspects, the intelligent tactile feedback, and human-machine interaction data. On this basis, a human-machine interaction optimization scheme during the telesurgery process is presented from four aspects, i.e., Edge-Cloud Integration, network slice, and intelligent edge-cloud. Finally, this paper discusses the open issues of the presented telesurgery system regarding the ultra-high reliability, AI-enabled surgery robot, communication, and security, to provide the reference for the promotion of the telesurgery robot performance.     

A conceptual framework for tactually guided exploration and shapeperception

A conceptual framework for tactually guided exploration and shape perception using a robotic medium is provided. The conceptual framework identifies the needed sensory information, the spatial and temporal transformations of this information, the control mechanism, both feedforward and feedback, for performing the missions and the perception machinery. These attributes, in turn, sharpen the focus on the major building blocks needed in design of artificial skins, tactile sensing, and processing systems of the future. These building blocks are identification of central nervous system (CNS) machinery in living systems that perform the required computations, mappings, processors for extraction of the needed manipulation and recognition parameters, processing of outputs of populations of natural tactile sensors, and finally, understanding the dynamics of sensor-imbedded skin and artificial skin in fixed, gliding, and rolling contact with known and unknown objects and surfaces     

A Motion Vector Sensor via Direct‐Current Triboelectric Nanogenerator

Motion vector sensors play an important role in artificial intelligence and internet of things. Here, a triboelectric vector sensor (TVS) based on a direct‐current triboelectric nanogenerator is reported, for self‐powered measuring various motion parameters, including displacement, velocity, acceleration, angular, and angular velocity. Based on the working mechanism of the contact‐electrification effect and electrostatic breakdown, a continuous DC signal can be collected to directly monitor moving objects free from environmental electromagnetic signal interference existing in conventional self‐powered TVSs with an alternative‐current signal output, which not only enhances the sensitivity of sensors, but also provides a simple solution to miniaturize the sensors. Its sensitivity is demonstrated to be equivalent to state‐of‐the‐art photoelectric technology by a comparative experiment in an intelligent mouse. Notably, an intelligent pen based on the miniaturized TVS is designed to realize motion trajectory tracing, mapping, and writing on the curved surface. This work provides a new paradigm shift to design motion vector sensors and self‐powered sensors in artificial intelligent and internet of things.     

Analyzing of flexible gripper by computational intelligence approach

Adaptive grippers should be able to detect and recognize grasping objects. To be able to do it control algorithm need to be established to control gripper tasks. Compliant underactuated mechanisms with passive behavior can be used for modelling of adaptive robotic fingers. Undearactuation is a feature which allows fully adaptability of robotic fingers for different objects. In this study gripper with two fingers was established. Finite element method (FEM) procedure was used to optimize the gripper structural topology. Kinetostatic model of the underactuated finger mechanism was analyzed. This design of the gripper has embedded sensors as part of its structure. The use of embedded sensors in a robot gripper gives the control system the ability to control input displacement of the gripper and to recognize specific shapes of the grasping objects. Since the conventional control strategy is a very challenging task, soft computing based controllers are considered as potential candidates for such an application. The sensors could be used for grasping shape detection. Given that the contact forces of the finger depend on contact position of the finger and object, it is suitable to make a prediction model for the contact forces in function of contact positions of the finger and grasping objects. The prediction of the contact forces was established by using a soft computing (computational intelligence) approach. To perform the contact forces estimation adaptive neuro-fuzzy (ANFIS) methodology was used. FEM simulations were performed in order to acquire experimental data for ANFIS training. The main goal was to apply ANFIS network in order to find correlation between sensors’ stresses and finger contact forces. Afterwards ANFIS results were compared with benchmark models (extreme learning machine (ELM), extreme learning machine with discrete wavelet algorithm (ELM-WAVELET), support vector machines (SVM), support vector machines with discrete wavelet algorithm (SVM-WAVELET), genetic programming (GP) and artificial neural network (ANN)). The reliability of these computational models was analyzed based on simulation results.     

Stretchable Energy‐Harvesting Tactile Interactive Interface with Liquid‐Metal‐Nanoparticle‐Based Electrodes

Energy‐harvesting electronic skin (E‐skin) is highly promising for sustainable and self‐powered interactive systems, wearable human health monitors, and intelligent robotics. Flexible/stretchable electrodes and robust energy‐harvesting components are critical in constructing soft, wearable, and energy‐autonomous E‐skin systems. A stretchable energy‐harvesting tactile interactive interface is demonstrated using liquid metal nanoparticles (LM‐NPs)‐based electrodes. This stretchable energy‐harvesting tactile interface relies on triboelectric nanogenerator composed of a galinstan LM‐NP‐based stretchable electrode and patterned elastic polymer friction and encapsulation layer. It provides stable and high open‐circuit voltage (268 V), short‐circuit current (12.06 µA), and transferred charges (103.59 nC), which are sufficient to drive commercial portable electronics. As a self‐powered tactile sensor, it presents satisfactory and repeatable sensitivity of 2.52 V·kPa^?1 and is capable of working as a touch interactive keyboard. The demonstrated stretchable and robust energy‐harvesting E‐skin using LM‐NP‐based electrodes is of great significance in sustainable human–machine interactive system, intelligent robotic skin, security tactile switches, etc.     

具有正则化约束的脉冲神经网络机器人触觉物体识别方法

拓展触觉感知能力是智能机器人未来发展的重要方向之一,决定着机器人的应用场景范围。由触觉传感器采集的数据是机器人完成触觉感知任务基础,但触觉数据具有复杂的时空性。脉冲神经网络具有丰富的时空动力学特征和契合硬件的事件驱动性,能更好地处理时空信息和应用于人工智能芯片给机器人带来更高能效。该文针对脉冲神经网络神经元脉冲活动离散性导致网络训练过程反向传播失效的问题,从智能触觉机器人动态系统角度,引入脉冲活动近似函数使脉冲神经网络反向传播梯度下降法有效;针对触觉脉冲数据量少导致的过拟合问题,融合正则化方法加以缓解;最后,提出具有正则化约束的脉冲神经网络机器人触觉物体识别(Spiking neural network Tactile dropout, SnnTd; Spiking neural network Tactile dropout-l2-cosine annealing, SnnTdlc)算法。相较于经典方法TactileSGNet, Grid-based CNN, MLP和GCN, SnnTd正则化方法触觉物体识别率在EvTouch-Containers数据集上比最好方法TactileSGNet提升了5.00%,SnnTdlc正则化方法触觉物体识别率在EvTouch-Objects数据集上比最好方法TactileSGNet提升了3.16%。     

浅谈人工智能在融媒体业务生产中的应用发展

近年来,随着IT技术和互联网的发展,人工智能AI(Artificial Intelligence)技术在广电行业得到广泛应用。文章基于融媒体环境下人工智能技术的应用与发展进行了探讨,分析人工智能技术如何推动生产效率提高以及智能机器人书写稿件、智能人脸识别、智能语音语义识别、智能化的OCR识别、自动播出等方面的应用与发展。     

Bioinspired Triboelectric Nanogenerators as Self‐Powered Electronic Skin for Robotic Tactile Sensing

Electronic skin (e‐skin) has been under the spotlight due to great potential for applications in robotics, human–machine interfaces, and healthcare. Meanwhile, triboelectric nanogenerators (TENGs) have been emerging as an effective approach to realize self‐powered e‐skin sensors. In this work, bioinspired TENGs as self‐powered e‐skin sensors are developed and their applications for robotic tactile sensing are also demonstrated. Through the facile replication of the surface morphology of natural plants, the interlocking microstructures are generated on tribo‐layers to enhance triboelectric effects. Along with the adoption of polytetrafluoroethylene (PTFE) tinny burrs on the microstructured tribo‐surface, the sensitivity for pressure measurement is boosted with a 14‐fold increase. The tactile sensing capability of the TENG e‐skin sensors are demonstrated through the characterizations of handshaking pressure and bending angles of each finger of a bionic hand during handshaking with human. The TENG e‐skin sensors can also be utilized for tactile object recognition to measure surface roughness and discern hardness. The facile fabrication scheme of the self‐powered TENG e‐skin sensors enables their great potential for applications in robotic dexterous manipulation, prosthetics, human–machine interfaces, etc.     

增强现实(AR)技术标准研究

未来二、三十年里,人类社会最大的变化是走向智能化的社会,未来身边所有的物体都会具有感知能力,而5G的一个重要使命就是让这些物体都能连接到网络上,然后通过网络后端的云计算,通过人工智能,来提供更好的服务。人类已经创再出能力超越自身的智能,而下一阶段逐步将人工智能应用于生活中,5G时代的大带宽、高速率、短时延就是为应用而生,而最有可能成为5G时代杀手级应用的则是增强现实(AR)技术。     

智能语音识别技术的发展现状与应用前景

在我们的日常生活中,语言是一个不可缺少的信息传播媒介,也是语音识别技术的亮点。近几年人工智能技术不断发展,已经应用到我们生活的各个领域,在目前的消费级市场中,语音交互是人工智能最普遍的表现方式。本论文对智能语音识别技术的发展做了简单介绍,并介绍了智能语音交互的概念,智能语音识别技术在智能家居中的应用,同时指出了其发展所存在的问题,并对未来的发展作出展望。     

Towards robust grasping: An analysis of in-hand object motion with FBG optical fibers as force sensing technology

To enable or allow for advanced manipulation scenarios in the envisaged future in robotic manipulation, the introduction of tactile or force-sensing technology is essential. This work introduces a study on a novel integration of fiber sensing elements inside the fingers of robotics grippers that allow obtaining accurate interaction forces during object manipulation. The study focuses on in-hand object motion and analyzes the objects’ dynamic behaviors during grasping or contact phases. The proposed approach classifies the behavior of the gripper-object interaction in real-time by looking at the measured forces in the orthogonal direction of the gripper’s fingers after a filtering stage. The methodology is validated theoretically and evaluated with real experiments using objects presenting different shapes, sizes, and surface properties.     

人工智能在智慧能源管理中的应用研究

物联网和云计算等技术的发展使得数据中心呈现爆炸性的增长,数据中心能耗大、能效利用率低、能耗和能效管理信息化程度不高,造成能源的巨大浪费。本文基于人工智能算法,从能耗指标PUE智能优化和能量智能管理调度角度,提供智慧能源管理解决方案。     

边缘智能在轨道交通中的应用：前景与展望

边缘智能作为一项新兴技术,正受到国内外学者的广泛关注,其作为人工智能技术与边缘计算技术的结合,有望促进人工智能技术在各行业的部署,加速产业智能化进程。该文首先介绍了边缘智能技术的基本原理、系统架构及其比较优势,梳理了边缘智能技术的国内外研究现状;分析了边缘智能在轨道交通建设工程、运维调度、智能控制、改造升级的全生命周期应用前景,详述了边缘智能技术在轨道交通过程管理控制、建设现场数据采集分析、信息共享、智能运维、智能调度、自动驾驶系统、列车协同控制及改造升级等全生命周期中的赋能作用。该文随后设计与实现了轨道交通智能运行控制为背景下的边缘智能平台,测试基于深度学习和强化学习的边缘智能应用的功能及性能。最后,归纳了边缘智能技术在轨道交通领域应用的问题与挑战。该文的研究期望为轨道交通领域的边缘智能应用提供有益的借鉴和实践基础。     

Development and Experimental Analysis of a Soft Compliant Tactile Microsensor for Anthropomorphic Artificial Hand

This paper presents the development and preliminary experimental analysis of a soft compliant tactile microsensor (SCTM) with minimum thickness of 2 mm. A high shear sensitive triaxial force microsensor was embedded in a soft, compliant, flexible packaging. The performance of the whole system, including the SCTM, an electronic hardware and a processing algorithm, was evaluated by static calibration, maximum load tests, noise and dynamic tests, and by focusing on slippage experiments. A proper tradeoff between final robustness and sensitivity of the tactile device was identified. The experiments showed that the tactile sensor is sufficiently robust for application in artificial hands while sensitive enough for slip event detection. The sensor signals were elaborated with the cumulative summation algorithm and the results showed that the SCTM system could detect a slip event with a delay from a minimum of 24.5 ms to a maximum of 44 ms in the majority of experiments fulfilling the neurophysiological requirement.