Nanyang Technological University, Singapore (NTU Singapore) is launching three new space projects under Singapore’s Space Technology Development Programme (STDP), marking a major step in accelerating the commercialisation of space technologies developed in Singapore. Among the projects is a satellite that will carry edge-computing artificial intelligence capable of processing data directly in space. Images captured by the satellite will be analysed on board using small AI models and an edge engine, reducing the need to transmit large volumes of raw data back to Earth and enabling faster, more intelligent decision-making in orbit.

To date, generative AI has struggled to create video based on music. While recent video models can produce visually impressive short clips, they tend to be less successful with long-form storytelling, musical alignment and character consistency.  

Now, a new system is using songs as the basis for generating complete music videos. 

Researchers at Koç University have developed an open-access, machine learning–assisted tool that automates microfluidic chip design, lowering the technical barrier for nonexpert users. Published in Science Advances and led by Assoc. Prof. Dr. Savaş Taşoğlu, the study introduces μFluidicGenius (μFG), a user-friendly platform that enables rapid design of microfluidic circuits based on target flow conditions, producing 3D-printable designs with approximately 90% accuracy.

Two-dimensional semiconductors are promising alternatives to silicon for future electronics, but controlling their atomic-scale growth remains a major challenge. Now, researchers from Okayama University, in collaboration with Shinshu University and Keio University, have directly observed in real time how monolayer transition metal dichalcogenides form in a confined microreactor. By visualizing molten precursor droplets during growth, the team identified distinct growth regimes determining crystal size and quality, offering strategies for designing high-performance semiconductor materials.