Researchers discovered the spin-configuration of excited states in a promising zero-dimensional luminescent material, (Bmpip)₂SnBr₄, resolved a scientific debate with groundbreaking insights into its unique energy structures and shed light on designing better LEDs and spin-based devices.

Recently, Hangzhou Institute of Medicine (HIM), University of Chinese Academy of Sciences (UCAS), Guangzhou National Laboratory and other research institutes published a perspective entitled Redefining imaging genomics for the next decade on Science Bulletin, in which the article systematically summarizes the existing advancements in imaging genomics, and proposes a framework for imaging genomics that provide a new path for precision medicine.

This study reviews how optimization techniques can improve the economic dispatch of local energy sources, helping reduce costs, enhance grid reliability, and support renewable integration. By comparing classical and heuristic methods, it identifies strategies that align with the UN Sustainable Development Goals (SDGs), particularly affordable and clean energy (SDG 7) and climate action (SDG 13), contributing to a more sustainable and efficient decentralized power system.

In recent years, the large-scale grid connection of various distributed power sources has made the planning and operation of distribution grids increasingly complex. Consequently, a large number of active distribution network reconfiguration techniques have emerged to reduce system losses, improve system safety, and enhance power quality via switching switches to change the system topology while ensuring the radial structure of the network. While scholars have previously reviewed these methods, they all have obvious shortcomings, such as a lack of systematic integration of methods, vague classification, lack of constructive suggestions for future study, etc. Therefore, this paper attempts to provide a comprehensive and profound review of 52 methods and applications of active distribution network reconfiguration through systematic method classification and enumeration. Specifically, these methods are classified into five categories, i.e., traditional methods, mathematical methods, meta-heuristic algorithms, machine learning methods, and hybrid methods. A thorough comparison of the various methods is also scored in terms of their practicality, complexity, number of switching actions, performance improvement, advantages, and disadvantages. Finally, four summaries and four future research prospects are presented. In summary, this paper aims to provide an up-to-date and well-rounded manual for subsequent researchers and scholars engaged in related fields.

A new study shows that treating mesenchymal stem cells with lipopolysaccharide (LPS) shifts their energy use toward glycolysis, reducing their healing potential in stroke therapy — offering insights into improving cell-based treatments for brain injuries.

Researchers from Tsinghua University and Henan Normal University developed a dedicated data analysis framework for the Silicon Strip Detector Telescopes (SSDTs) of the Compact Spectrometer for Heavy-IoN Experiments (CSHINE). Based on a modular architecture design, the framework integrates core analysis steps—including detector calibration, particle identification, and track reconstruction—into a unified system through C++ classes, effectively addressing the technical challenges of processing complex SSDT signals. Its robust performance was demonstrated through the successful analysis of light-charged particles in the 25 MeV/u ⁸⁶Kr + ¹²⁴Sn experiment conducted at the first Radioactive Ion Beam Line in Lanzhou, allowing for precise extraction of physical observables, including energy, momentum, and particle type. Through the optimization of track recognition algorithms with the utilization of reconstructed physical data, including effective physical event counts and energy spectra, the research team significantly enhanced the track recognition efficiency, achieving a remarkable rate of approximately 90%. This framework provides a standardized and reusable technical solution for SSDT-based detector systems.

Researchers at Tianjin University have developed LineGen, a physics-guided method combining ordinary differential equations and a time-embedded U-Net, to generate super-resolution load data (SRLD) for power distribution systems. Published in Engineering, the approach addresses limitations in high-frequency data collection by achieving a 1000-fold resolution enhancement—significantly surpassing prior deep learning methods—while offering traceable physical interpretability for improved grid management and reliability.

Researchers at Shenzhen University have developed a novel self-sensing steel fiber-reinforced polymer  composite bar (SFCB) that integrates distributed fiber-optic sensors (DFOS) for real-time monitoring of cracks and mechanical behavior in reinforced concrete members. This innovative approach, detailed in a recent study published in Engineering, aims to enhance the durability and safety of civil infrastructure by providing a more accurate and reliable method for structural health monitoring.