What is heat, and what is useful work if a machine consists only of an atom and light particles? In modern quantum technologies, this kind of question connects thermodynamics with quantum physics. Researchers at the University of Basel, Switzerland, have developed a theoretical approach that can reconcile both theories.  

Designing Latent Heat Thermal Energy Storage (LHTES) systems requires time-consuming and computationally expensive computational fluid dynamics simulations. To overcome this bottleneck, a new study presents a hybrid physics-informed neural network (PINN) framework trained on a sparse ground-truth dataset. Together with a genetic algorithm-based multi-objective design optimization, this PINN framework enables rapid, autonomous exploration of the continuous LHTES design space and thermohydraulic trade-offs. This approach serves as a crucial tool for developing next-generation LHTES systems.

For sleep-deprived nursing moms, artificial intelligence may feel like a lifeline. In fact, a new national survey from Orlando Health reveals if women ages 18 to 44 had an urgent breastfeeding question or concern half (51%) of them would first turn to an AI chatbot, social media or a search engine.