Bidirectional visual-tactile cross-modal generation using latent feature space flow model.

Inspired by visual-tactile cross-modal bidirectional mapping of the human brain, this paper introduces a novel approach to bidirectional mapping between visual and tactile data, an area not fully explored in the predominantly unidirectional existing studies. First, we adopt separate Variational AutoEncoder (VAE) models for visual and tactile data. Furthermore, we introduce a conditional flow model built on the VAE latent feature space, enabling cross-modal bidirectional mapping between visual and tactile data using one model. The experimental results show that our method achieves excellent performance in terms of the similarity between the generated data and the original data (Structural Similarity Index (SSIM) of visual data: 0.58, SSIM of tactile data: 0.80), the classification accuracy on generated data (visual data: 91.60%, tactile data: 88.05%), and the zero-shot classification accuracy between generated data and language (visual data: 44.49%, tactile data: 45.03%). To the best of our knowledge, the method proposed in this paper is the first one to utilize a single model to achieve bidirectional mapping between visual and tactile data. Our model and code will be made public after the acceptance of the paper.

mRNA-based Vaccines Targeting the T-cell Epitope-rich Domain of Epstein Barr Virus Latent Proteins Elicit Robust Anti-Tumor Immunity in Mice.

Epstein-Barr virus (EBV) is associated with various malignancies and infects >90% of the global population. EBV latent proteins are expressed in numerous EBV-associated cancers and contribute to carcinogenesis, making them critical therapeutic targets for these cancers. Thus, this study aims to develop mRNA-based therapeutic vaccines that express the T-cell-epitope-rich domain of truncated latent proteins of EBV, including truncatedlatent membrane protein 2A (Trunc-LMP2A), truncated EBV nuclear antigen 1 (Trunc-EBNA1), and Trunc-EBNA3A. The vaccines effectively activate both cellular and humoral immunity in mice and show promising results in suppressing tumor progression and improving survival time in tumor-bearing mice. Furthermore, it is observed that the truncated forms of the antigens, Trunc-LMP2A, Trunc-EBNA1, and Trunc-EBNA3A, are more effective than full-length antigens in activating antigen-specific immune responses. In summary, the findings demonstrate the effectiveness of mRNA-based therapeutic vaccines targeting the T-cell-epitope-rich domain of EBV latent proteins and providing new treatment options for EBV-associated cancers.

Direct recycling of spent Li-ion batteries: Challenges and opportunities toward practical applications.

With the exponential expansion of electric vehicles (EVs), the disposal of Li-ion batteries (LIBs) is poised to increase significantly in the coming years. Effective recycling of these batteries is essential to address environmental concerns and tap into their economic value. Direct recycling has recently emerged as a promising solution at the laboratory level, offering significant environmental benefits and economic viability compared to pyrometallurgical and hydrometallurgical recycling methods. However, its commercialization has not been realized in the terms of financial feasibility. This perspective provides a comprehensive analysis of the obstacles that impede the practical implementation of direct recycling, ranging from disassembling, sorting, and separation to technological limitations. Furthermore, potential solutions are suggested to tackle these challenges in the short term. The need for long-term, collaborative endeavors among manufacturers, battery producers, and recycling companies is outlined to advance fully automated recycling of spent LIBs. Lastly, a smart direct recycling framework is proposed to achieve the full life cycle sustainability of LIBs.

Direct Regeneration of NCM Cathode Material with Aluminum Scraps.

Hydrothermal-based direct regeneration of spent Li-ion battery (LIB) cathodes has garnered tremendous attention for its simplicity and scalability. However, it is heavily reliant on manual disassembly to ensure the high purity of degraded cathode powders, and the quality of regenerated materials. In reality, degraded cathodes often contain residual components of the battery, such as binders, current collectors, and graphite particles. Thorough investigation is thus required to understand the effects of these impurities on hydrothermal-based direct regeneration. In this study, we focus on isolating the effects of aluminum (Al) scraps on the direct regeneration process. We found that Al metal can be dissolved during the hydrothermal relithiation process. Even when the cathode material contains up to 15 wt.% Al scraps, no detrimental effects were observed on the recovered structure, chemical composition, and electrochemical performance of the regenerated cathode material. The regenerated NCM cathode can achieve a capacity of 163.68 mAh/g at 0.1 C and exhibited a high-capacity retention of 85.58 % after cycling for 200 cycles at 0.5 C. Therefore, the hydrothermal-based regeneration method is effective in revitalizing degraded cathode materials, even in the presence of notable Al impurity content, showing great potential for industrial applications.

Design and research of soft-body cavity-type detonation drivers.

According to the high-energy-density movement characteristics of animals during jumping, soft-body cavity-type detonation driver that combines the explosive chemical reaction mechanism of hydrogen and oxygen is designed, in order to control the robot in jump to achieve output optimization. Then, combined with the theoretical values of the detonation dynamic equation and experimental data for the performance parameters, the influences of the mixing ratio of hydrogen (H2) and oxygen (O2), the volume of mixed hydrogen and oxygen in the cavity, and the shape, wall thickness, and area ratio value of the soft-body cavity on the output performance of the detonation driver are analyzed. When gas volume is 20:10 mL, the jump height reaches 2.5 m. In addition, the upper and lower area ratio of cavity is optimized to 2:1, improving the output performance by 21.6% on average. Therefore, the above research results provide reference for the driver structure design of jumping robot.

Miniature Amphibious Robot Actuated by Rigid-Flexible Hybrid Vibration Modules.

Amphibious robots can undertake various tasks in terrestrial and aquatic environments for their superior environmental compatibility. However, the existing amphibious robots usually utilize multi-locomotion systems with transmission mechanisms, leading to complex and bulky structures. Here, a miniature amphibious robot based on vibration-driven locomotion mechanism is developed. The robot has two unique rigid-flexible hybrid modules (RFH-modules), in which a soft foot and a flexible fin are arranged on a rigid leg to conduct vibrations from an eccentric motor to the environment. Then, it can run on ground with the soft foot adopting the friction locomotion mechanism and swim on water with the flexible fin utilizing the vibration-induced flow mechanism. The robot is untethered with a compact size of 75 × 95 × 21 mm3 and a small weight of 35 g owing to no transmission mechanism or joints. It realizes the maximum speed of 815 mm s-1 on ground and 171 mm s-1 on water. The robot, actuated by the RFH-modules based on vibration-driven locomotion mechanism, exhibits the merits of miniature structure and fast movements, indicating its great potential for applications in narrow amphibious environments.

A Task-Learning Strategy for Robotic Assembly Tasks from Human Demonstrations.

In manufacturing, traditional task pre-programming methods limit the efficiency of human-robot skill transfer. This paper proposes a novel task-learning strategy, enabling robots to learn skills from human demonstrations flexibly and generalize skills under new task situations. Specifically, we establish a markerless vision capture system to acquire continuous human hand movements and develop a threshold-based heuristic segmentation algorithm to segment the complete movements into different movement primitives (MPs) which encode human hand movements with task-oriented models. For movement primitive learning, we adopt a Gaussian mixture model and Gaussian mixture regression (GMM-GMR) to extract the optimal trajectory encapsulating sufficient human features and utilize dynamical movement primitives (DMPs) to learn for trajectory generalization. In addition, we propose an improved visuo-spatial skill learning (VSL) algorithm to learn goal configurations concerning spatial relationships between task-relevant objects. Only one multioperation demonstration is required for learning, and robots can generalize goal configurations under new task situations following the task execution order from demonstration. A series of peg-in-hole experiments demonstrate that the proposed task-learning strategy can obtain exact pick-and-place points and generate smooth human-like trajectories, verifying the effectiveness of the proposed strategy.

Robot complex motion learning based on unsupervised trajectory segmentation and movement primitives.

This paper presents a robot skill acquisition framework for learning and reproducing humanoid trajectories with complex forms. A new unsupervised segmentation method is proposed to detect motion units in the demonstrated kinematic data using the concept of key points. To find the consistent features of trajectories, a Hidden Semi-Markov Model (HSMM) is used to identify key points common to all the demonstrations. Generalizing the motion units is achieved via a Probability-based Movement Primitive (PbMP), which encapsulates multiple trajectories into one model. Such a framework can generate trajectories suitable for robot execution with arbitrary shape and complexity from a small number of demonstrations, which greatly expands the application scenarios of robot programming by demonstration. The automatic segmentation process does not rely on a priori knowledge or models for specific tasks, and the generalized trajectory retains more consistent features than those produced by other algorithms. We demonstrate the effectiveness of the proposed framework through simulations and experiments.

Periodontal regeneration using bone morphogenetic protein 2 incorporated biomimetic calcium phosphate in conjunction with barrier membrane: A pre-clinical study in dogs.

AIM: To evaluate the effect of bone morphogenetic protein 2 (BMP-2) incorporated biomimetic calcium phosphate (BMP-2/BioCaP) in conjunction with barrier membrane on periodontal regeneration in chronic periodontitis experimental model. MATERIAL AND METHODS: Chronic periodontitis experimental model with critical-sized supra-alveolar defects was created in 15 dogs' mandibles. After the initial periodontal therapy, the defects were randomly assigned to the following groups: (a) control; (b) barrier membrane; (c) deproteinized bovine bone mineral + barrier membrane; (d) BioCaP + barrier membrane and (e) BMP-2/BioCaP + barrier membrane (6 quadrants with 18 teeth per group). Eight weeks later, clinical examinations, micro-CT, and histomorphometric analyses were performed. RESULTS: Clinical examinations, including plaque index, bleeding index, and probing depth, were similar for all groups. In contrast, the clinical attachment loss was significantly lower in defects grafted with BMP-2/BioCaP and barrier membrane. The micro-CT results showed that the height of mineralized tissue in defects grafted with BMP-2/BioCaP and barrier membrane was significantly higher. For histometric analysis, the defects grafted with BMP-2/BioCaP and barrier membrane exhibited significantly more connective tissue height, new cementum height, new bone height and area, as well as less down-growth of junctional epithelium. CONCLUSION: BMP-2/BioCaP could be a promising bone substitute for periodontal regeneration.

Monocular Vision-Based Pose Determination in Close Proximity for Low Impact Docking.

Pose determination in close proximity is critical for space missions in which monocular vision is one of the most promising solutions. Although numerous approaches such as using artificial beacons or specific shapes on spacecrafts have proved to be effective, the high individuation and the large time delay limit their use in low impact docking. This paper proposes a unified framework to determinate the relative pose between two docking mechanisms by treating their guide petals as measurement objects. Fusing the pose information of one docking mechanism to simplify image processing and creating an intermediate coordinate system to solve the perspective-n-point problem greatly improve the real-time performance and the robustness of the method. Experimental results show that the position measurement error is within 3.7 mm, while the rotation error around docking direction is less than 0.16°, corresponding to a measurement time reduction of 85%.

Biomechanical design of escalading lower limb exoskeleton with novel linkage joints.

In this paper, an obstacle-surmounting-enabled lower limb exoskeleton with novel linkage joints that perfectly mimicked human motions was proposed. Currently, most lower exoskeletons that use linear actuators have a direct connection between the wearer and the controlled part. Compared to the existing joints, the novel linkage joint not only fitted better into compact chasis, but also provided greater torque when the joint was at a large bend angle. As a result, it extended the angle range of joint peak torque output. With any given power, torque was prioritized over rotational speed, because instead of rotational speed, sufficiency of torque is the premise for most joint actions. With insufficient torque, the exoskeleton will be a burden instead of enhancement to its wearer. With optimized distribution of torque among the joints, the novel linkage method may contribute to easier exoskeleton movements.

A water-walking robot mimicking the jumping abilities of water striders.

The highly efficient and agile water-surface locomotion of water striders has attracted substantial research attention. Compared with imitating the horizontal rowing motion, imitating the jumping capability of water striders is much more challenging because the strong interaction in the jumping process easily causes the robot to sink. This study focuses on designing a miniature robot capable of continuously jumping on the water surface. A spring-based actuating mechanism is proposed to produce a large jumping force. The center of gravity of the robot is carefully designed to allow the robot to jump on the surface continuously and smoothly. The influences of several critical factors, including the area of the supporting legs, the spring stiffness, the jumping angle, etc on jumping ability are analyzed by means of dynamic simulation and experiments. The jumping performance under different jumping angles is tested. The fabricated robot weighs approximately 10.2 g and can continuously jump on water with a maximum leap height and length of 120 and 410 mm, respectively. This study helps researchers understand the jumping mechanism of water striders and provides a reference for developing water-jumping robots that can perform various aquatic tasks in the future.

Biomechanical modeling and load-carrying simulation of lower limb exoskeleton.

This paper introduces novel modern equipment-a lower extremity exoskeleton, which can implement the mutual complement and the interaction between human intelligence and the robot's mechanical strength. In order to provide a reference for the exoskeleton structure and the drive unit, the human biomechanics were modeled and analyzed by LifeModeler and Adams software to derive each joint kinematic parameter. The control was designed to implement the zero-force interaction between human and exoskeleton. Furthermore, simulations were performed to verify the control and assist effect. In conclusion, the system scheme of lower extremity exoskeleton is demonstrated to be feasible.

Research on musculoskeletal model of elbow joint for evaluating the feasibility of FES.

Tremor usually occurs in a patient's upper limb with a roughly sinusoidal profile. Understanding the inner mechanism of the involuntary movement is fundamental to improving tremor suppression treatments. Therefore, the musculoskeletal model of the elbow joint was developed in this study. Initially, healthy subjects were selected to simulate tremor and the tremulous data was collected with the purpose of sparing patients from fatigue. With the recorded joint angle and surface EMG (sEMG), the model was calibrated to subjects by optimization approach. The activation derived from the electric pulse was employed to drive the tuned model and the model's output was compared with the angle predicted by the EMG-driven musculoskeletal model. The results demonstrated that the performance of the calibrated model was improved by a smaller normalized root mean square error and a higher coefficient of determination compared with the no-tuned model. There was no significant difference between the angles estimated by the tuned model activated by the electric pulse and muscle excitation. It indicates that neural activation could be replaced by the electric pulse to excite the limbs for desired angle. Therefore, the study presents a good way to evaluate the feasibility of Functional Electric Stimulation to suppress tremor.