Search

Large volumes of used electronics are often collected in remanufacturing plants, which requires disassembly before harvesting parts for reuse. Disassembly is mainly conducted manually with low productivity. Recently, human–robot collaboration has been considered as a solution. To assist effectively, robots should observe work environments and recognize human actions accurately. Rich activity video recording and supervised learning can be used to extract insights; however, supervised learning does not allow robots to self-accomplish the learning process. This study proposes an unsupervised learning framework for achieving video-based human activity recognition. The framework consists of two main elements: 1) a variational-autoencoder-based architecture for unlabeled data representation learning and 2) a hidden Markov model for activity state division. The complete explicit activity classification is validated against ground truth labels; here, we use a case study of disassembling a hard disk drive. The framework shows an average recognition accuracy of 91.52%, higher than competing methods.

The three-dimensional structure of chromatin plays a crucial role in development and disease, both of which are associated with transcriptional changes. However, given the heterogeneity in single-cell chromatin architecture and transcription, the regulatory relationship between the three-dimensional chromatin structure and gene expression is difficult to explain based on bulk cell populations. Here we develop a single-cell, multimodal, omics method allowing the simultaneous detection of chromatin architecture and messenger RNA expression by sequencing (single-cell transcriptome sequencing (scCARE-seq)). Applying scCARE-seq to examine chromatin architecture and transcription from 2i to serum single mouse embryonic stem cells, we observe improved separation of cell clusters compared with single-cell chromatin conformation capture. In addition, after defining the cell-cycle phase of each cell through chromatin architecture extracted by scCARE-seq, we find that periodic changes in chromatin architecture occur in parallel with transcription during the cell cycle. These findings highlight the potential of scCARE-seq to facilitate comprehensive analyses that may boost our understanding of chromatin architecture and transcription in the same single cell.

Artificial nanoenzymes with enzyme-like catalytic activity have gradually become an alternative to natural enzymes due to their low production cost, high stability, and good tolerance. In recent years, various enzyme mimics have emerged with the rapid development of nano-teclnology. Metal-organic frameworks(MOFs) are a novel class of porous inorganic-organic hybrid materials made from metal ions/clusters and organic ligands, and MOFs-based nanozymes show great prospect in biosensing, biocatalysis, biomedical imaging, and therapeutic applications, due to unique properties, such as high specific surface area, high porosity, tunable morphology, and excellent biocatalytic properties. In this paper, the recent progresses concerning MOFs-based nanozymes are systematically summarized, including the synthesis, design strategies and related applications, which are divided into two major categories, namely, MOFs structured nanoenzymes and MOFs composite structured nanoenzymes. Meanwhile, the applications of various classifications of MOFs research are introduced. At the end, current challenges and future perspectives of MOFs-based nanozymes are also discussed. It is highly expected that this review on this important area can provide a meaningful guidance for tumor therapy, biosensing and other aspects.

Human intention prediction plays a critical role in human–robot collaboration, as it helps robots improve efficiency and safety by accurately anticipating human intentions and proactively assisting with tasks. While current applications often focus on predicting intent once human action is completed, recognizing human intent in advance has received less attention. This study aims to equip robots with the capability to forecast human intent before completing an action, i.e., early intent prediction. To achieve this objective, we first extract features from human motion trajectories by analyzing changes in human joint distances. These features are then utilized in a Hidden Markov Model (HMM) to determine the state transition times from uncertain intent to certain intent. Second, we propose two models including a Transformer and a Bi-LSTM for classifying motion intentions. Then, we design a human–robot collaboration experiment in which the operator reaches multiple targets while the robot moves continuously following a predetermined path. The data collected through the experiment were divided into two groups: full-length data and partial data before state transitions detected by the HMM. Finally, the effectiveness of the suggested framework for predicting intentions is assessed using two different datasets, particularly in a scenario when motion trajectories are similar but underlying intentions vary. The results indicate that using partial data prior to the motion completion yields better accuracy compared to using full-length data. Specifically, the transformer model exhibits a 2% improvement in accuracy, while the Bi-LSTM model demonstrates a 6% increase in accuracy.

As a result of increasingly stringent power supply requirements and diverse application scenarios, the development of hybrid nanogenerators that can combine multiple power generation modes has gradually become a hot research topic. In this work, the PR-HNG, a piezo/triboelectric hybrid nanogenerator based on a pitched roof structure, is proposed and is used efficiently to perform discrete mechanical energy harvesting. Using the flexible connection of its separated roof-like trigger mechanism, the PR-HNG operates through inertia under vibration conditions and achieves a satisfactory working state with very low mechanical loss. At 10 Hz, the instantaneous open-circuit voltage VOCand short-circuit current ISCof the PR-HNG can reach 800 V and 9.6 mA (1.60 A/m2), respectively, and no obvious attenuation occurs during high-intensity and long-term operation. After connection to a load via a rectifier bridge, the PR-HNG has a maximum power output of 575 μW/cm2and can charge a 2 mF capacitor up to more than 2.5 V in only 30 s; these two performance parameters are better than the reported values for other similar types of HNG. In practical applications, this PR-HNG can easily drive more than 1440 commercial light-emitting diodes and can drive small electronic devices in real time. Therefore, the PR-HNG has been demonstrated to provide a credible and promising solution for fabrication of self-powered monitoring systems for use in industrial environments.

•Iron overload aggravates the difficulty of engraftment of transplanted stem cells.•Mesenchymal stem cells (MSCs) aid hematopoietic stem cell (HSC) engraftment.•We study the effect of MSC cotransplantation on cell engraftment in iron overload mice.•MSC cotransplantation improved the engraftment of umbilical cord blood mononuclear cells (UCB-MNCs) in iron overload mice.•MSC cotransplantation improved the proliferation of UCB-MNCs in iron overload mice.