Sensory neurons within skin form an interface between the external physical reality and the inner tactile perception. This interface enables sensory information to be organized identified, and interpreted through perceptual learning-the process whereby the sensing abilities improve through experience. Here, an artificial sensory neuron that can integrate and differentiate the spatiotemporal features of touched patterns for recognition is shown. The system comprises sensing, transmitting, and processing components that are parallel to those found in a sensory neuron. A resistive pressure sensor converts pressure stimuli into electric signals, which are transmitted to a synaptic transistor through interfacial ionic/electronic coupling via a soft ionic conductor. Furthermore, the recognition error rate can be dramatically decreased from 44% to 0.4% by integrating with the machine learning method. This work represents a step toward the design and use of neuromorphic electronic skin with artificial intelligence for robotics and prosthetics.

X. (2020). Gesture recognition using a bioinspired learning architecture that integrates visual data with somatosensory data from stretchable sensors.

Numerical simulations on fluid dynamics problems primarily rely on spatially or/and temporally discretization of the governing equation using polynomials into a finite-dimensional algebraic system. Due to the multi-scale nature of the physics and sensitivity from meshing a complicated geometry, such process can be computational prohibitive for most realtime applications (e.g., clinical diagnosis and surgery planning) and many-query analyses (e.g., optimization design and uncertainty quantification). Therefore, developing a costeffective surrogate model is of great practical significance. Deep learning (DL) has shown new promises for surrogate modeling due to its capability of handling strong nonlinearity and high dimensionality. However, the off-the-shelf DL architectures, success of which heav-* Corresponding author. of internal flows relevant to hemodynamics applications, and the forward propagation of uncertainties in fluid properties and domain geometry is studied as well. The results show excellent agreement on the flow field and forward-propagated uncertainties between the DL surrogate approximations and the first-principle numerical simulations.

Mechanical responsiveness in many plants is produced by helical organizations of cellulose microfibrils. However, simple mimicry of these naturally occurring helical structures does not produce artificial materials with the desired tunable actuations. Here, we show that actuating fibres that respond to solvent and vapour stimuli can be created through the hierarchical and helical assembly of aligned carbon nanotubes. Primary fibres consisting of helical assemblies of multiwalled carbon nanotubes are twisted together to form the helical actuating fibres. The nanoscale gaps between the nanotubes and micrometre-scale gaps among the primary fibres contribute to the rapid response and large actuation stroke of the actuating fibres. The compact coils allow the actuating fibre to rotate reversibly. We show that these fibres, which are lightweight, flexible and strong, are suitable for a variety of applications such as energy-harvesting generators, deformable sensing springs and smart textiles.

interact with the world. This inspires scientists and engineers to develop flexible and stretchable electronic devices or systems to emulate the functionality of the human skin, known as electronic skin (e-skin). Recently, tremendous efforts have been made in the development of e-skin from materials innovation to structural designs, which concentrate on the improvement of sensing capability (i.e., stretchability, sensitivity, long-term monitoring, etc.), user-friendly detection mode (i.e., noninvasive, inflammation-free, gaspermeable, implantable, etc.), systemlevel integration (i.e., data transmission, power supply, etc.), and the realization of new functions (i.e., self-healing). [5][6][7] However, as one of the essential elements to mimic human intelligence, the functionality of perception, that is, interpretation of acquired sensory data, is still lacking in most e-skin systems. Implementing the perception functionality in a flexible and stretchable sensing system, referred to as artificial skin perception is vital to realizing an authentically intelligent artificial skin, with capabilities beyond human skin (Figure 1b). Coupled with sensing, feedback, and other technologies (Figure 1c), artificial skin perception will significantly accelerate the development of next-generation soft robotics, where a low-latency and energy-efficient data processing module is required to enable fast adaptation to dynamic environments.Currently, perception processes of most e-skin systems take place in centralized processing units, that is, computers or servers in the cloud, located far away from the sensing systems where sensory signals are generated. These sensory signals, usually time-serial, unstructured, and redundant, need to be continually sent to an external processing end. As a result, there will be a tremendous volume of data movements between the sensing end and the processing end, leading to huge energy consumption. [8,9] Moreover, the frequent and continual exchange of data causes a serious burden to data communication due to the limited bandwidth of communication channels in current sensing systems, especially when a large number of sensors are mounted on the large-area sensing system. [9][10][11] This leads to a notorious latency problem, known as a time delay in response for data communication. [9] The latency issue severely hinders the development of ultrafast responsive and delaysensitive intelligent systems, such as advanced robotics and Skin is the largest organ, with the functionalities of protection, regulation, and sensation. The emulation of human skin via flexible and stretchable electronics gives rise to electronic skin (e-skin), which has realized artificial sensation and other functions that cannot be achieved by conventional electronics. To date, tremendous progress has been made in data acquisition and transmission for e-skin systems, while the implementation of perception within systems, that is, sensory data processing, is still in its infancy. Integrating the perception functionality into a flex...