Ultrasonic testing is a promising non-destructive evaluation technique across various industries. In a novel breakthrough, researchers from Chung-Ang University have developed DiffectNet, an AI-based technology that facilitates the diffusion-enabled conditional target generation of internal defects in ultrasonic non-destructive testing. This approach significantly outperforms traditional methods, potentially revolutionizing real-time defect reconstruction and prediction in highly reliability-critical industries, including aerospace, power generation, semiconductor manufacturing, and civil infrastructure.

Last week, four lasers were projected into the skies above the European Southern Observatory’s (ESO’s) Paranal site in Chile. The lasers are each used to create an artificial star, which astronomers use to measure and then correct the blur caused by Earth's atmosphere. The striking launch of these lasers, one from each of the eight-metre telescopes at Paranal, is a significant milestone of the GRAVITY+ project — a large and complex upgrade to ESO’s Very Large Telescope Interferometer (VLTI). GRAVITY+ unlocks a greater observing power and a much wider sky coverage for the VLTI than previously possible.

The recent test, announced by the European Southern Observatory (ESO) today, mark a significant milestone of the GRAVITY+ project, a complex upgrade to ESO’s Very Large Telescope Interferometer (VLTI), unlocking greater observing power and much wider sky coverage.

The malignant pediatric brain tumor medulloblastoma is driven by proteins deep within the cancer cells, for which there are often no effective drugs available. Researchers at Uppsala University have identified a new genetic technique that enables targeting of tumour cells with elevated levels of such proteins. The technique could be used to eliminate cells responsible for tumour relapse. The hope is to develop an effective drug against recurrence of this disease within a few years.

X-ray absorption spectroscopy (XAS) provides valuable information about a material’s properties and electronic states. However, it requires extensive expertise and manual effort for conventional analysis. Now, researchers from Japan have developed a novel artificial intelligence-based approach for analyzing XAS data that can enable rapid, autonomous, and object material identification. This novel approach outperforms the previous studies in terms of higher accuracy, accelerating the development of new materials.