Delivering high-definition television over local 5G networks could help modernize cable TV in buildings where installing fiber-optic cables is impractical. However, since lost data cannot be retransmitted in 5G multicast broadcasting, temporary video freezing is common. Now, researchers from Japan have developed a lightweight AI model that predicts wireless conditions in advance and proactively adjusts transmission settings, enabling stable, real-time video delivery via 5G broadcasting. Their work helps bridge the information gap in underserved areas. 

Distributed sensor networks use data from multiple sensors to locate targets. However, tracking many targets at once requires sorting out which measurements belong to which targets across sensors, which is difficult without a centralized system. To address this problem, researchers from China have developed MASTER, a distributed optimization method that combines swarm optimization with contribution-based cooperation. Experiments show that MASTER achieves lower localization errors and more stable consensus than existing distributed methods.

Accurate and efficient isolation and recovery of interest cells allow detailed analyses of cellular function, growth and culture, and gene-expression profiling of cells within heterogeneous populations. Researchers have developed a novel microfluidic surface-enhanced Raman scattering (SERS) chip capable of performing deep-learning-assisted, label-free single-cell identification and sorting. The results could be used to guide future research on the design and development of non-invasive, efficient cell-sorting techniques with potential applications in clinical diagnostics and precision medicine.

Nanoporous carbon materials have attracted significant interest for applications in lightweight structural materials, microdevices, and bio-integrated systems due to their low density, chemical stability, electrical conductivity, and tunable mechanical properties. However, existing fabrication approaches often struggle to simultaneously achieve precise patterning, controllable porosity, and desirable mechanical performance. A research team led by Professor Min Qiu at Westlake University has developed a new fabrication strategy, termed Mix-ice lithography (Mix-IL), that enables direct construction of nanoporous carbon structures with tunable pore architectures and complex geometries. The study has been published in Science Bulletin.

Researchers from Shanxi University and Nanjing University have demonstrated on-chip two-mode squeezed light at audio frequency as low as 60 Hz, bringing integrated quantum noise reduction into the audio-frequency band. A weak, orthogonally polarized electro-optic reference comb stabilizes the homodyne measurement phase without placing a bright locking tone at the quantum frequencies. The experiment observed approximately 1 dB of squeezing from 60 Hz to 5 kHz, reconstructed the two-mode covariance matrix, and verified entanglement using the positive partial transposition criterion.

The central north Amazon, with extensive areas of high forest cover, natural savannas and vast and important indigenous territories, was not previously considered as being the most affected by climate change.

However, this region, which is considered critical to Amazonia’s cultural and biodiversity wealth has unexpectedly found itself at the forefront of rapid growth in climate extremes, a new study reveals.

Findings, published today in Communications Earth & Environment, identify Amazonia as experiencing rapid growth in extreme temperatures and water stress – with 10% of the basin, an area larger than 700,000 square kilometres (bigger than Afghanistan), witnessing increases in extreme temperatures during the dry season of at least 0.75°C a decade and more than 3.22°C since 1981.

A nationwide survey in Japan suggests cultural worldviews are more strongly linked to support for climate policies than to support for biodiversity conservation measures. While hierarchical worldviews were linked to lower perceived climate risks and weaker support for climate measures, these associations were weaker for climate-related biodiversity conservation approaches and not clearly observed for stand-alone conservation policies, pointing to potential common ground.

Associate Professor Kenta Kasai of Science Tokyo has developed a high-rate quantum low-density parity-check (LDPC) code by adapting design theory from classical communications and storage. By imposing orthogonality only where quantum mechanics requires it, the method retains the freedom used to control connection degrees, randomness, and short loops in classical LDPC codes. Overseas teams have already extended the design to neutral-atom hardware and broader code frameworks, referring to the resulting family as “Kasai codes.”

Eight technologists building five Singapore-based deep-tech start-ups have been selected for the inaugural cohort of Activate Global Fellows – Singapore, a national fellowship hosted by Nanyang Technological University, Singapore (NTU Singapore) to support promising science and technology ventures develop into globally competitive companies.