A deep learning-based model for estimating pollution fluxes from rivers into the sea and its optimization.

Pollution fluxes from rivers into the sea are currently the main source of pollutants in nearshore areas. Based on the source-sink process of the basin-estuary-coastal waters system, the pollution fluxes into the sea and their spatiotemporal heterogeneity were estimated. A deep learning-based model was established to simplify the estimation of pollution fluxes into the sea, with socio-economic drivers and meteorological data as input variables. A method for estimating the contribution rate of pollution fluxes from different spatial gradient was proposed. In this study, we found that (1) the pollution fluxes into the sea of total nitrogen (TN) and total phosphorus (TP) from the Bohai Sea Rim Basin (BSRB) in 1980, 1990, 2000, 2010, and 2020 were 25.38 × 104, 26.12 × 104, 27.27 × 104, 29.82 × 104, 25.31 × 104 and 1.32 × 104, 2.14 × 104, 2.09 × 104, 1.87 × 104, 1.68 × 104 tons, respectively. (2) The proportion of rural life and livestock to the TN was the highest, accounting for 39.18 % and 21.19 %, respectively. The proportion of livestock to the TP was the highest, accounting for 39.20 %, followed by rural life, accounting for 24.72 %. The results indicated that the pollution fluxes in the BSRB were related to human economic activities and relevant environmental protection measures. (3) The deep learning-based model established to estimate runoff pollution fluxes into the sea had the accuracy of over 90 %. (4) As for contribution rate, in terms of the elevation, the range of 0-100 m had the highest proportion, accounting for 39.65 %. The range of 50-100 km from the coastline had the highest proportion, accounting for 18.11 %. In terms of the district, coastal area has the highest proportion, accounting for 38.00 %. This study revealed the changing trends and driving mechanisms of pollution fluxes into the sea over the past 40 years and established a simplified deep learning-based model for estimating pollution fluxes into the sea. Then, we identified regions with high pollution contribution rate. The results can provide scientific references for the adaptive management of the nearshore areas based on the ecosystem.

Marangoni-driven deterministic formation of softer, hollow microstructures for sensitivity-enhanced tactile system.

Microengineering the dielectric layers with three-dimensional microstructures has proven effective in enhancing the sensitivity of flexible pressure sensors. However, the widely employed geometrical designs of solid microstructures exhibit limited sensitivity over a wide range of pressures due to their inherent but undesired structural compressibility. Here, a Marangoni-driven deterministic formation approach is proposed for fabricating hollow microstructures, allowing for greater deformation while retarding structural stiffening during compression. Fluid convective deposition enables solute particles to reassemble in template microstructures, controlling the interior cavity with a void ratio exceeding 90%. The hollow micro-pyramid sensor exhibits a 10-fold sensitivity improvement across wider pressure ranges over the pressure sensor utilizing solid micro-pyramids, and an ultra-low detect limit of 0.21 Pa. With the advantages of facilitation, scalability, and large-area compatibility, such an approach for hollow microstructures can be expanded to other sensor types for superior performance and has considerable potential in robotic tactile and epidermal devices.

Unleashing the Potential of MXene-Based Flexible Materials for High-Performance Energy Storage Devices.

Since the initial discovery of Ti3 C2 a decade ago, there has been a significant surge of interest in 2D MXenes and MXene-based composites. This can be attributed to the remarkable intrinsic properties exhibited by MXenes, including metallic conductivity, abundant functional groups, unique layered microstructure, and the ability to control interlayer spacing. These properties contribute to the exceptional electrical and mechanical performance of MXenes, rendering them highly suitable for implementation as candidate materials in flexible and wearable energy storage devices. Recently, a substantial number of novel research has been dedicated to exploring MXene-based flexible materials with diverse functionalities and specifically designed structures, aiming to enhance the efficiency of energy storage systems. In this review, a comprehensive overview of the synthesis and fabrication strategies employed in the development of these diverse MXene-based materials is provided. Furthermore, an in-depth analysis of the energy storage applications exhibited by these innovative flexible materials, encompassing supercapacitors, Li-ion batteries, Li-S batteries, and other potential avenues, is conducted. In addition to presenting the current state of the field, the challenges encountered in the implementation of MXene-based flexible materials are also highlighted and insights are provided into future research directions and prospects.

A Snakeskin-Inspired, Soft-Hinge Kirigami Metamaterial for Self-Adaptive Conformal Electronic Armor.

A majority of soft-body creatures evolve armor or shells to protect themselves. Similar protection demand is for flexible electronics working in complex environments. Existing works mainly focus on improving the sensing capabilities such as electronic skin (E-skin). Inspired by snakeskin, a novel electronic armor (E-armor) is proposed, which not only possesses mechanical flexibility and electronic functions similar to E-skin, but is also able to protect itself and the underlying soft body from external physical damage. The geometry of the kirigami mechanical metamaterial (Kiri-MM) ensures auxetic stretchability and meanwhile large areal coverage for sufficient protection. Moreover, to suppress the inherent but undesired out-of-plane buckling of conventional Kiri-MMs for conformal applications, soft hinges are used to form a distinct soft (hinges)-rigid (tiles) configuration. Analytical, computational, and experimental studies of the mechanical behaviors of the soft-hinge Kiri-MM E-armor demonstrate the merits of this design, i.e., stretchability, conformability, and protectability, as applied to flexible electronics. Deploying a conductive soft material at the hinges enables facile wiring strategies for large-scale circuit arrays. Functional E-armor systems for controllable display and sensing purposes provide simple examples of a wide spectrum of applications of this concept.

Accuracy of predicting metabolizable energy from in vitro digestible energy determined with a computer-controlled simulated digestion system in feed ingredients for ducks.

Two experiments were conducted to study the accuracy of predicting true metabolizable energy (TME) of ingredients for ducks from in vitro digestible energy (IVDE) determined with a computer-controlled simulated digestion system. Experiment 1 was to establish TME prediction models from the IVDE of 9 energy feed ingredients and 12 protein feed ingredients using regression analysis. Experiment 2 was to validate the accuracy of the predicted ME of 10 ingredients randomly selected from Exp. 1. Ten diets were formulated with 2 to 6 of 10 ingredients. Dietary in vivo TME values were compared with calculated values based on the TME predicted in Exp. 1. In Exp. 1, the correlation coefficients between TME and IVDE were 0.9339 (P < 0.05) in 9 energy feed ingredients and 0.8332 (P < 0.05) in 12 protein feed ingredients. No significant difference was observed on the slope and intercept of TME regression models between 9 energy feed ingredients and 12 protein feed ingredients. Therefore, the regression model of TME on IVDE for 21 feed ingredients was TME = 0.7169 × IVDE +1,224 (R 2 = 0.7542, P < 0.01). Determined and predicted TME differed by less than 100 kcal/kg of DM in 11 ingredients, and the difference ranged from 100 to 200 kcal/kg of DM in 5 ingredients. However, the difference between determined and predicted TME varied from 410 to 625 kcal/kg of DM in rice bran, rapeseed meal, corn gluten meal, and citric acid meal. In Exp. 2, the determined and calculated TME were comparable (3,631 vs. 3,639 kcal/kg of DM) and highly correlated (r = 0.9014; P < 0.05) in 10 diets. Determined and calculated TME differed by less than 100 kcal/kg of DM in 7 diets and by 106 to 133 kcal/kg of DM in 3 diets. These results have demonstrated that TME can be accurately predicted from IVDE in most feed ingredients, but it is less accurate for rice bran, rapeseed meal, corn gluten and citric acid meal.

Highly Robust and Wearable Facial Expression Recognition via Deep-Learning-Assisted, Soft Epidermal Electronics.

The facial expressions are a mirror of the elusive emotion hidden in the mind, and thus, capturing expressions is a crucial way of merging the inward world and virtual world. However, typical facial expression recognition (FER) systems are restricted by environments where faces must be clearly seen for computer vision, or rigid devices that are not suitable for the time-dynamic, curvilinear faces. Here, we present a robust, highly wearable FER system that is based on deep-learning-assisted, soft epidermal electronics. The epidermal electronics that can fully conform on faces enable high-fidelity biosignal acquisition without hindering spontaneous facial expressions, releasing the constraint of movement, space, and light. The deep learning method can significantly enhance the recognition accuracy of facial expression types and intensities based on a small sample. The proposed wearable FER system is superior for wide applicability and high accuracy. The FER system is suitable for the individual and shows essential robustness to different light, occlusion, and various face poses. It is totally different from but complementary to the computer vision technology that is merely suitable for simultaneous FER of multiple individuals in a specific place. This wearable FER system is successfully applied to human-avatar emotion interaction and verbal communication disambiguation in a real-life environment, enabling promising human-computer interaction applications.

Evaluation system of coastal wetland ecological vulnerability under the synergetic influence of land and sea: A case study in the Yellow River Delta, China.

A comprehensive evaluation system and model of Coastal Wetland Ecological Vulnerability (CWEV) was constructed and applied to reveal spatial heterogeneity of the ecological vulnerability of the Yellow River Delta Wetland (YRDW). The results showed that the score of the ecological vulnerability (EVS) of the YRDW was 0.49, which was generally at a medium vulnerability level. The wetland area of high vulnerability was up to 943km2, accounting for 35.2% of the total area, followed by the medium vulnerable area with an area of 750km2, accounting for 28.1% of the total area. From the coastline perpendicularly to the land, the "seaward" gradient effect gradually decreased, the vulnerability-increasing "hydrologic connectivity" effect increased with the distance from the river channel, and the "land source influence" effect gradually decayed along with the vulnerability of population and economy gathering areas.