Rapid and accurate detection of tomatoes in natural plantation environments is the linchpin for smart harvesting. However, the complexity of existing models necessitates high hardware computing power, constraining the deployment of tomato recognition models in picking robots. In this study, the overall scheme framework of the vision algorithm is divided into three stages. In the first stage, based on YOLOv8s, a small object detection layer was added along with hyperparameter optimization, resulting in the proposed YOLOv8-TomatoTeacher (YOLOv8-TT) to improve the accuracy and precise positioning of tomato detection, laying the foundation for subsequent operations. In the second stage, the pruning algorithm is applied to the trained YOLOv8-TT model to trim and optimize redundant channels, yielding the pruned lightweight YOLOv8-TomatoStudent (YOLOv8-TS) model. In the third stage, using YOLOv8-TT as the teacher model and YOLOv8-TS as the student model, the detection accuracy of YOLOv8-TS was further enhanced through knowledge distillation, thus obtaining the YOLOv8-TomatoDetector (YOLOv8-TD) that considers both detection speed and accuracy. Experimental results indicate that YOLOv8-TD demonstrates significant advantages in parameter quantity and model complexity, with only 0.35 M Params and 4.4G GFlops, which are respectively reduced by 96.86 % and 84.51 % compared with baseline model YOLOv8s. In addition, YOLOv8-TD achieved a 1.4 % increase in mAP, and the detection speed of 354f/s satisfies real-time requirements. The model proposed in this study can swiftly and accurately detect dense small tomato targets, reducing the deployment cost and difficulty of the robot vision system.

To prevent mango fruits from being damaged in mango fruit robotic picking, accurate and cost-effective simultaneous detection of mango fruits and fruiting stems is a key upstream task. Yet accomplishing such task remains a notable challenge due to complex and resource-limited orchard environments such as power supply, occlusion, variable light and color similarity. To address these issues, this study proposed an improved YOLOv8 model and successfully deployed it on an edge device, enabling a simultaneous detection of mango fruits and fruiting stem. Specifically, a BRA sparse attention module combined with transformer is firstly introduced into the backbone network of YOLOv8, to reduce the interference of orchard green background on mango detection. Then, the detection head of YOLOv8 was replaced via dynamic detection head combined with deformable convolution operator, to improve the detection precision for elongated mango fruiting stems. Finally, the feature fusion network of YOLOv8 is reconstructed by using GSConv-based Slimneck structure to achieve lightweight neck network, thus facilitating subsequent deployment on edge device. Experimental results showed that the improved YOLOv8 model achieved a detection precision of 97.63 % in mango fruits and a detection precision 94.5 % in fruiting stems, both surpassing lightweight state-of-the-art models including YOLOv5n, YOLOv5s, YOLOv5lite, YOLOv7tiny and YOLOV8s. Moreover, compared with the original YOLOv8n model, the improved YOLOv8 model increased the detection precision of mango fruits by 1.43 %, and the detection precision of fruiting stems by 2.8 %, while maintaining the lightweight model characteristic. To verify the effectiveness of the improved YOLOv8 model on the edge device, the proposed model was deployed on NVIDIA Jetson Orin Nano to achieve effective simultaneous detection of mango fruits and fruiting stems. This study can offer valuable technological support for automatic mango fruit robotic picking.

Positioning of mango picking points is a crucial technology for the realisation of automated robotic mango harvesting. Herein, this study reported a visualised end-to-end system for mango picking point positioning using improved YOLOv8 architecture with object detection and instance segmentation, as well as an algorithm of picking point positioning. At first, the improved YOLOv8n model, incorporating the BiFPN structure and the SPD-Conv module, was utilised to enhance the detection performance of mango fruits and stems. This model achieved a detection precision of 98.9% in fruits and 97.1% in stems, with recall of 99.5% and 94.6% respectively. Then, the YOLOv8n-seg model was used for segment the stem ROI (Region of interest), leading to 81.85% in MIoU and 88.69% in mPA. Finally, a skeleton line of the stem region was obtained on the basis of the segmentation image, and a picking point positioning algorithm was developed to determine the coordinates of the optimal picking point. Subsequently, the positioning success rate of coordinates, absolute errors, and relative errors were calculated by comparing the automatic positioned coordinates with the manually positioned stem region. Experimental results indicated that this study achieved an average positioning success rate of 92.01%, with an average absolute error of 4.93 pixels and an average relative error of 13.11%. Additionally, the average processing time for processing 640 images using the picking point positioning system is 72.75 ms. This study demonstrates the reliability and effectiveness of positioning mango picking points, laying the technological basis for the automated harvesting of mango fruits.

Microchannel-type flexible strain sensors have aroused great interest in the fields of medical devices, artificial skin, and soft robotics due to their high stretchability, low hysteresis, and excellent durability. However, they are often limited in practical applications because of low sensitivity. In this article, a novel liquid metal-based sensor with nonuniform strain distribution was developed, aimed at enhancing the sensor sensitivity. The sensor was designed with localized changes in cross-sectional area using 3-D printed molds. Compared with the traditional microchannel-type flexible strain sensors with uniform strain distribution, the sensitivity of our developed sensor increased by 18%, demonstrating a good improvement. In addition, the sensor exhibits low hysteresis (0.59%), fast response and recovery time (184 ms), long durability (1000 cycles of stretching/releasing at 100% strain), and excellent dynamic and static performance. Finally, the sensor was applied to a robotic joint and human body joints, such as the neck, elbow, knee, and wrist to capture various activities. The results indicate that the developed flexible strain sensor exhibits outstanding performance, making it a promising candidate for wearable devices.

Accurate remaining useful life (RUL) prediction of rolling bearings plays a vital role in ensuring the safe operation of mechanical equipment. Graph-based models have become an emerging trend in RUL prediction by converting monitoring samples into graph structures to capture samples’ relationships effectively. However, graph-based models only use pairwise samples to model the relationships between samples and cannot capture the non-pairwise high-order relationships between multiple samples. Besides, graph-based models rely heavily on predefined graphs to aggregate relevant features. The bearing monitoring datasets have no explicit structure, and the predefined graph structures cannot characterize datasets. Aiming at these issues, a temporal multi-resolution hypergraph attention network (T-MHGAT) is proposed. Firstly, the bearings’ monitoring samples are established and fused into a multi-resolution hypergraph (MHG) to characterize the potential structure of bearings monitoring datasets. Then, a hypergraph attention network (HGAT) is designed to mine the high-order relationships between signal samples on hypergraph data. Meanwhile, multiple gated recurrent units (GRUs) are constructed to capture the signal samples’ temporal information. Finally, the linear layer is built after GRUs to output RUL prediction values. Many experiments on two rolling bearing datasets showed the effectiveness of T-MHGAT, which can lay the foundation for predictive equipment maintenance.

Precise detection and low-cost deployment are the technological basis of intelligent fruit picking. This study proposes a lightweight improved YOLOv5s model to detect pitaya fruits in daytime and nighttime lightsupplement environments, and make it successfully deploy in an Android device. This model first uses the module of shufflenetv2 to reconstruct the YOLOv5s backbone network. Then, the study proposes a ConcentratedComprehensive Convolution Receptive Field Enhancement (C3RFE) module to improve the detection precision of pitaya fruits. Furthermore, a Bidirectional Feature Pyramid Network (BiFPN) feature fusion method is used to enhance the multi-scale feature fusion. Moreover, three optimized Squeeze-and-Excitation (SE) attention modules are added to make full use of the image feature information. Finally, a dynamic label allocation strategy simple Optimal Transport Assignment (simOTA) is used to optimize the YOLOv5s model original label allocation strategy. The experimental results show that the improved model achieves an average precision rate of 97.80 %, with frames per second (FPS) of 139 FPS in a GPU run environment. The model size is only 2.5 MB. Compared to the state-of-the-art SSD, Faster RCNN, YOLOv4, YOLOv4 tiny, YOLOv5s, YOLOv5Lite-s, YOLOXs, YOLOv7, YOLOv7-tiny, YOLOv8n and YOLOv8s, the improved YOLOv5s achieve preferred comprehensive performance in average precision rate, FPS and model size. When deploying this model on the Realme GT Android mobile phone by developing an application based on an NCNN framework, such an application accomplishes real-time pitaya fruit detection with an FPS exceeding 30 FPS. This study can provide technological support for a precise and effective pitaya fruit intelligent picking.

Precise positioning of fruit and picking point is crucial for harvesting table grapes using automated picking robots in an unstructured agricultural environment. Most studies employ multi-step methods for locating picking points based on fruit detection, leading to slow detection speed, cumbersome models, and algorithmic fragmentation. This study proposes an improved YOLOv8-GP (YOLOv8-Grape and picking point) model based on YOLOv8n-Pose to solve the problem of simultaneous detection of grape clusters and picking points. YOLOv8-GP is an end-to-end network that integrates object detection and key point detection. Specifically, the Bottleneck in C2f is replaced with FasterNet Block that incorporates EMA (Efficient Multi-Scale Attention), resulting in C2f-Faster-EMA. BiFPN is applied to substitute the original PAN as Neck network. The FasterNet Block, designed based on partial convolution (PConv), reduces redundant computation and memory access, thereby extracting spatial features more efficiently. The EMA attention mechanism achieves performance gains with lower computational overhead. Furthermore, BiFPN is employed to enhance the effect of feature fusion. Experimental results demonstrate that YOLOv8-GP achieves AP of 89.7 % for grape cluster detection and a Euclidean distance error of less than 30 pixels for picking point detection. Additionally, the number of Params is reduced by 47.73 %, and the model complexity GFlops is 6.1G. In summary, YOLOv8-GP offers excellent detection performance, while the reduced number of parameters and model complexity contribute to lower deployment costs and easier implementation on mobile robots.

Sugarcane field re-seeding robot is a promising yield-enhancing technology proposed to solve the seedling absences in sugarcane fields. In this study, in combination with developing the sugarcane field re-seeding robot, an improved YOLOv5s model was proposed to detect sugarcane seedlings and predict seed replenishment positions. Firstly, field images of one-month-old sugarcane seedlings were taken at different light conditions as a dataset. Secondly, the Slim-Neck was introduced to replace the Neck network, which can reduce the complexity of the model while maintaining sufficient accuracy. Thirdly, the Efficient Channel Attention (ECA) module was added to the Backbone network to enhance the model's attention on critical feature information of sugarcane seedlings. Fourthly, the SCYLLA-IoU (SIoU) loss function was introduced to speed up the convergence of the proposed model. Lastly, a method for predicting seed replenishment positions was proposed and verified by the field tests. The experimental results showed that the mean average precision (mAP), precision, and recall of the improved YOLOv5s model were 93.1 %, 92.1 %, and 89.9 %, respectively, and the detection speed was 82 frames per second (FPS), which increased the mAP by 1.5 % and the detection speed by 12.3 % compared to the original YOLOv5s model. In addition, compared with Faster R-CNN, SSD, and YOLOv4-tiny models, the improved YOLOv5s model had a higher accuracy, faster detection speed, and less memory consumption. The field test showed that the real-time detection speed of the improved YOLOv5s model was 23 FPS in Nvidia Jetson TX2. The real-time detection precision of sugarcane seedlings was 97.2 %, and the recall was 86.7 %. The mean relative error between the numbers of seed replenishment positions predicted by the robot and that predicted by the human was 18.7 %. Consequently, the improved YOLOv5s model can efficiently and accurately detect sugarcane seedlings and predict seed replenishment positions. This technology provides valuable visual detection support for the sugarcane field re-seeding robot.

The extraction of high-quality RNA from kenaf is essential for genetic and molecular biology research. However, the presence of high levels of polysaccharide and polyphenol compounds in kenaf poses challenges for RNA isolation. We proposed a simplified, time-saving and cost-effective method for isolating high quantities of RNA from various kenaf tissues. This method exhibited superior efficiency in RNA isolation compared with the conventional cetyltrimethylammonium bromide method and demonstrated greater adaptability to different samples than commercial kits. Furthermore, the high-quality RNA obtained from this method was successfully utilized for RT-PCR, real-time RT-PCR and northern blot analysis. Moreover, this proposed protocol also enables the acquisition of both high-quality and -quantity gDNA through RNase A treatment. In addition, the effectiveness of this approach in isolating high-quality RNA from other plant species has been experimentally confirmed.

DNA methylation is defined as a conserved epigenetic modification mechanism that plays a key role in maintaining normal gene expression without altering the DNA sequence. Several studies have reported that altered methylation patterns were associated with male sterility in some plants such as rice and wheat, but global methylation profiles and their possible roles in cytoplasmic male sterility (CMS), especially in cotton near-isogenic lines, remain unclear. In this study, bisulfite sequencing technology and RNA-Seq were used to investigate CMS line 07-113A and its near-isogenic line 07-113B. Using integrated methylome and transcriptome analyses, we found that the number of hypermethylated genes in the differentially methylated regions, whether in the promoter region or in the gene region, was more in 07-113A than the number in 07-113B. The data indicated that 07-113A was more susceptible to methylation. In order to further analyze the regulatory network of male sterility, transcriptome sequencing and DNA methylation group data were used to compare the characteristics of near-isogenic lines 07-113A and 07-113B in cotton during the abortion stage. Combined methylation and transcriptome analysis showed that differentially expressed methylated genes were mainly concentrated in vital metabolic pathways including the starch and sucrose metabolism pathways and galactose metabolism. And there was a negative correlation between gene methylation and gene expression. In addition, five key genes that may be associated with CMS in cotton were identified. These data will support further understanding of the effect of DNA methylation on gene expression and their potential roles in cotton CMS.

d-Allulose is a desirable sucrose substitute with potential applications in food and health care. d-Allulose can be synthesized using d-glucose as a substrate through coupling glucose isomerase with d-allulose 3-epimerase (DAEase); however, the product yield is typically less than 20% at reaction equilibrium and thus limits its use in industrial applications. Here, a 3R-ketose phosphorylation pathway coupled with an adenosine triphosphate (ATP) regeneration system was developed for the efficient synthesis of d-allulose in Escherichia coli using d-glucose as a substrate. The l-rhamnulose kinase (RhaB) was used to break the inherent reaction equilibrium due to its substrate specificity, resulting in increases in d-allulose titer by 69.9% to 4.96 ± 0.49 g/L. By optimizing the whole cell transformation conditions and designing an ATP regeneration module, d-allulose production reached 17.62 ± 0.77 g/L from 30 g/L d-glucose with a final yield of 0.73 g/g without the addition of exogenous ATP. To evaluate the potential industrial application of this multienzyme cascade system, d-allulose was produced from cane molasses (124.16 ± 2.69 g/L glucose equivalent) with a final d-allulose titer of 62.60 ± 3.76 g/L. The present study provides a practical enzymatic approach for the economical synthesis of d-allulose.

Gene expression and translation in plant mitochondria remain poorly understood due to the complicated transcription of its mRNA. In this study, we report the 5' and 3' RNA extremities and promoters of five mitochondrial genes, atp1, atp4, atp6, atp9, and cox3. The results reveal that four genes (atp1, atp4, atp6, and cox3) are transcribed from multiple initiation sites but with a uniform transcript at the 3' end, indicating that heterogeneity of the 5' end is a common feature in the transcription of kenaf mitochondrial genes. Furthermore, we found that the transcription initiation sites of these four genes are significantly different in UG93A, UG93B, and the F1 hybrid. These data indicate that nuclear loci and unknown transcription factors within the mitochondria of different cytoplasmic types may be involved in mitochondrial transcription. Promoter architecture analysis showed that the promoter core sequences are conserved in the kenaf mitochondrial genome but are highly divergent, suggesting that these elements are essential for the promoter activity of mitochondrial genes in kenaf. Our results reveal that the heterogeneity of the 5' end and uniformity at the 3' end are common transcriptional features of mitochondrial genes. These data provide essential information for understanding the transcription of mitochondrial genes in kenaf and can be used as a reference for other plants.