Large-Area and Low-Cost Force/Tactile Capacitive Sensor for Soft Robotic Applications

A Soft Robot Tactile Finger Using Oxidation-Reduction Graphene-Polyurethane Conductive Sponge

Currently, intelligent robotics is supplanting traditional industrial applications. It extends to business, service and care industries, and other fields. Stable robot grasping is a necessary prerequisite for all kinds of complex application scenarios. Herein, we propose a method for preparing an elastic porous material with adjustable conductivity, hardness, and elastic modulus. Based on this, we design a soft robot tactile fingertip that is gentle, highly sensitive, and has an adjustable range. It has excellent sensitivity (~1.089 kpa-1), fast response time (~35 ms), and measures minimum pressures up to 0.02 N and stability over 500 cycles. The baseline capacitance of a sensor of the same size can be increased by a factor of 5-6, and graphene adheres better to polyurethane sponge and has good shock absorption. In addition, we demonstrated the application of the tactile fingertip to a two-finger manipulator to achieve stable grasping. In this paper, we demonstrate the great potential of the soft robot tactile finger in the field of adaptive grasping for intelligent robots.

Compliant Grasping Control for a Tactile Self-Sensing Soft Gripper

Soft grippers with good passive compliance can effectively adapt to the shape of a target object and have better safe grasping performance than rigid grippers. However, for soft or fragile objects, passive compliance is insufficient to prevent grippers from crushing the target. Thus, to complete nondestructive grasping tasks, precision force sensing and control are immensely important for soft grippers. In this article, we proposed an online learning self-tuning nonlinearity impedance controller for a tactile self-sensing two-finger soft gripper so that its grasping force can be controlled accurately. For the soft gripper, its grasping force is sensed by a liquid lens-based optical tactile sensing unit that contains a self-sensing fingertip and a liquid lens module and has many advantages of a rapid response time (about 0.04 s), stable output, good sensitivity (>0.4985 V/N), resolution (0.03 N), linearity (R2 > 0.96), and low cost (power consumption: 5 mW, preparation cost Collapse

Key Words

• force control
• full online sequential extreme learning machine
• liquid lens-based optical tactile sensing unit
• soft gripper

Collapse

MESH Headings Collapse

Grants Collapse

Affiliation(s)

Hui Yang Dynamic Image Perception Lab, Institute of Semiconductors, Guangdong Academy of Sciences, Guangdong, China Biomechanics and Soft Robotics Lab, School of Mechanical Engineering and Automation, Beihang University, Beijing, China
Jiaqi Liu Biomechanics and Soft Robotics Lab, School of Mechanical Engineering and Automation, Beihang University, Beijing, China
Wenbo Liu Biomechanics and Soft Robotics Lab, School of Mechanical Engineering and Automation, Beihang University, Beijing, China
Weirui Liu Department of Mechanical and Electrical Engineering, School of Mechanical Engineering and Automation, Liaoning Petrochemical University, Fushun, China
Zilong Deng Department of Mechanical and Electrical Engineering, School of Mechanical Engineering and Automation, Liaoning Petrochemical University, Fushun, China
Yunzhi Ling Dynamic Image Perception Lab, Institute of Semiconductors, Guangdong Academy of Sciences, Guangdong, China
Changan Wang Dynamic Image Perception Lab, Institute of Semiconductors, Guangdong Academy of Sciences, Guangdong, China
Meixia Wu Dynamic Image Perception Lab, Institute of Semiconductors, Guangdong Academy of Sciences, Guangdong, China
Lihui Wang Dynamic Image Perception Lab, Institute of Semiconductors, Guangdong Academy of Sciences, Guangdong, China
Li Wen Biomechanics and Soft Robotics Lab, School of Mechanical Engineering and Automation, Beihang University, Beijing, China

Collapse

Active learning strategies for robotic tactile texture recognition tasks

Accurate texture classification empowers robots to improve their perception and comprehension of the environment, enabling informed decision-making and appropriate responses to diverse materials and surfaces. Still, there are challenges for texture classification regarding the vast amount of time series data generated from robots' sensors. For instance, robots are anticipated to leverage human feedback during interactions with the environment, particularly in cases of misclassification or uncertainty. With the diversity of objects and textures in daily activities, Active Learning (AL) can be employed to minimize the number of samples the robot needs to request from humans, streamlining the learning process. In the present work, we use AL to select the most informative samples for annotation, thus reducing the human labeling effort required to achieve high performance for classifying textures. We also use a sliding window strategy for extracting features from the sensor's time series used in our experiments. Our multi-class dataset (e.g., 12 textures) challenges traditional AL strategies since standard techniques cannot control the number of instances per class selected to be labeled. Therefore, we propose a novel class-balancing instance selection algorithm that we integrate with standard AL strategies. Moreover, we evaluate the effect of sliding windows of two-time intervals (3 and 6 s) on our AL Strategies. Finally, we analyze in our experiments the performance of AL strategies, with and without the balancing algorithm, regarding f1-score, and positive effects are observed in terms of performance when using our proposed data pipeline. Our results show that the training data can be reduced to 70% using an AL strategy regardless of the machine learning model and reach, and in many cases, surpass a baseline performance. Finally, exploring the textures with a 6-s window achieves the best performance, and using either Extra Trees produces an average f1-score of 90.21% in the texture classification data set.

Implementation of Field-Programmable Gate Array Platform for Object Classification Tasks Using Spike-Based Backpropagated Deep Convolutional Spiking Neural Networks

This paper investigates the performance of deep convolutional spiking neural networks (DCSNNs) trained using spike-based backpropagation techniques. Specifically, the study examined temporal spike sequence learning via backpropagation (TSSL-BP) and surrogate gradient descent via backpropagation (SGD-BP) as effective techniques for training DCSNNs on the field programmable gate array (FPGA) platform for object classification tasks. The primary objective of this experimental study was twofold: (i) to determine the most effective backpropagation technique, TSSL-BP or SGD-BP, for deeper spiking neural networks (SNNs) with convolution filters across various datasets; and (ii) to assess the feasibility of deploying DCSNNs trained using backpropagation techniques on low-power FPGA for inference, considering potential configuration adjustments and power requirements. The aforementioned objectives will assist in informing researchers and companies in this field regarding the limitations and unique perspectives of deploying DCSNNs on low-power FPGA devices. The study contributions have three main aspects: (i) the design of a low-power FPGA board featuring a deployable DCSNN chip suitable for object classification tasks; (ii) the inference of TSSL-BP and SGD-BP models with novel network architectures on the FPGA board for object classification tasks; and (iii) a comparative evaluation of the selected spike-based backpropagation techniques and the object classification performance of DCSNNs across multiple metrics using both public (MNIST, CIFAR10, KITTI) and private (INHA_ADAS, INHA_KLP) datasets.

Development of MWCNT/Magnetite Flexible Triboelectric Sensors by Magnetic Patterning

The fabrication of low-electrical-percolation-threshold polymer composites aims to reduce the weight fraction of the conductive nanomaterial necessary to achieve a given level of electrical resistivity of the composite. The present work aimed at preparing composites based on multiwalled carbon nanotubes (MWCNTs) and magnetite particles in a polyurethane (PU) matrix to study the effect on the electrical resistance of electrodes produced under magnetic fields. Composites with 1 wt.% of MWCNT, 1 wt.% of magnetite and combinations of both were prepared and analysed. The hybrid composites combined MWCNTs and magnetite at the weight ratios of 1:1; 1:1/6; 1:1/12; and 1:1/24. The results showed that MWCNTs were responsible for the electrical conductivity of the composites since the composites with 1 wt.% magnetite were non-conductive. Combining magnetite particles with MWCNTs reduces the electrical resistance of the composite. SQUID analysis showed that MWCNTs simultaneously exhibit ferromagnetism and diamagnetism, ferromagnetism being dominant at lower magnetic fields and diamagnetism being dominant at higher fields. Conversely, magnetite particles present a ferromagnetic response much stronger than MWCNTs. Finally, optical microscopy (OM) and X-ray micro computed tomography (micro CT) identified the interaction between particles and their location inside the composite. In conclusion, the combination of magnetite and MWCNTs in a polymer composite allows for the control of the location of these particles using an external magnetic field, decreasing the electrical resistance of the electrodes produced. By adding 1 wt.% of magnetite to 1 wt.% of MWCNT (1:1), the electric resistance of the composites decreased from 9 × 10⁴ to 5 × 10³ Ω. This approach significantly improved the reproducibility of the electrode's fabrication process, enabling the development of a triboelectric sensor using a polyurethane (PU) composite and silicone rubber (SR). Finally, the method's bearing was demonstrated by developing an automated robotic soft grip with tendon-driven actuation controlled by the triboelectric sensor. The results indicate that magnetic patterning is a versatile and low-cost approach to manufacturing sensors for soft robotics.

Multi-Transduction-Mechanism Technology, an Emerging Approach to Enhance Sensor Performance

Conventional sensor systems employ single-transduction technology where they respond to an input stimulus and transduce the measured parameter into a readable output signal. As such, the technology can only provide limited corresponding data of the detected parameters due to relying on a single transformed output signal for information acquisition. This limitation commonly results in the need for utilizing sensor array technology to detect targeted parameters in complex environments. Multi-transduction-mechanism technology, on the other hand, may combine more than one transduction mechanism into a single structure. By employing this technology, sensors can be designed to simultaneously distinguish between different input signals from complex environments for greater degrees of freedom. This allows a multi-parameter response, which results in an increased range of detection and improved signal-to-noise ratio. In addition, utilizing a multi-transduction-mechanism approach can achieve miniaturization by reducing the number of required sensors in an array, providing further miniaturization and enhanced performance. This paper introduces the concept of multi-transduction-mechanism technology by exploring different candidate combinations of fundamental transduction mechanisms such as piezoresistive, piezoelectric, triboelectric, capacitive, and inductive mechanisms.

Effects of Sensing Tactile Arrays, Shear Force, and Proprioception of Robot on Texture Recognition

In robotics, tactile perception is important for fine control using robot grippers and hands. To effectively incorporate tactile perception in robots, it is essential to understand how humans use mechanoreceptors and proprioceptors to perceive texture. Thus, our study aimed to investigate the impact of tactile sensor arrays, shear force, and the positional information of the robot's end effector on its ability to recognize texture. A deep learning network was employed to classify tactile data from 24 different textures that were explored by a robot. The input values of the deep learning network were modified based on variations in the number of channels of the tactile signal, the arrangement of the tactile sensor, the presence or absence of shear force, and the positional information of the robot. By comparing the accuracy of texture recognition, our analysis revealed that tactile sensor arrays more accurately recognized the texture compared to a single tactile sensor. The utilization of shear force and positional information of the robot resulted in an improved accuracy of texture recognition when using a single tactile sensor. Furthermore, an equal number of sensors placed in a vertical arrangement led to a more accurate distinction of textures during exploration when compared to sensors placed in a horizontal arrangement. The results of this study indicate that the implementation of a tactile sensor array should be prioritized over a single sensor for enhanced accuracy in tactile sensing, and the use of integrated data should be considered for single tactile sensing.

Development of Piezoelectric Silk Sensors Doped with Graphene for Biosensing by Near-Field Electrospinning

A novel piezoelectric fiber sensor based on polyvinylidene fluoride piezoelectric (PVDF) doped with graphene is presented. The near-field electrospinning technology was used for developing the sensor. The uniform experimental design method was introduced to determine the ranges of experimental parameters, including the applied voltage, the drum speed range, the graphene doping ratios from 0% to 11 wt% in PVDF solution, and the electrode gap. By experimental results, the conductivities of PVDF solutions with different doping ratios of graphene increased from 19.6 μS/cm to 115.8 μS/cm. Tapping tests were performed to measure the voltages and currents produced by the piezoelectric fibers. The maximum output voltage was 4.56 V at 5 wt% graphene doping ratio in PVDF fibers, which was 11.54 times that of the pure PVDF sensors. Moreover, mechanical properties of the proposed sensor were measured. Motion intention and swallowing test, such as saliva-swallowing and eating, were carried out. When the subject spoke normally, the output voltage of the sensor was between 0.2 and 0.4 V, approximately. Furthermore, when the subject drank water and ate food, the output voltage of the sensor was between 0.5 and 1 V, approximately. The proposed sensor could be used to detect signals of the human body and serve as a wearable device, allowing for more diagnosis and medical treatment.

Robot-Based Calibration Procedure for Graphene Electronic Skin

The paper describes the semi-automatised calibration procedure of an electronic skin comprising screen-printed graphene-based sensors intended to be used for robotic applications. The variability of sensitivity and load characteristics among sensors makes the practical use of the e-skin extremely difficult. As the number of active elements forming the e-skin increases, this problem becomes more significant. The article describes the calibration procedure of multiple e-skin array sensors whose parameters are not homogeneous. We describe how an industrial robot equipped with a reference force sensor can be used to automatise the e-skin calibration procedure. The proposed methodology facilitates, speeds up, and increases the repeatability of the e-skin calibration. Finally, for the chosen example of a nonhomogeneous sensor matrix, we provide details of the data preprocessing, the sensor modelling process, and a discussion of the obtained results.