When seawater penetrates concrete in road tunnels, a biofilm is formed that breaks down the concrete. This can lead to high costs and the risk of damage if stone and concrete drop from the tunnel roof.  In a new study, led by a team of researchers at Chalmers University of Technology in Sweden, research reveals the mechanisms behind the degradation, and its unexpectedly rapid progression.

To address the growing health threats posed by climate change, Nanyang Technological University, Singapore (NTU Singapore) is launching a new interdisciplinary research centre focused on climate change and environmental health in the tropics.

The use of AI-generated testimonials with increasingly realistic human features is becoming widespread on social media platforms like TikTok, even when it comes to communicating scientific content. It’s now quite common to come across videos where icons like Albert Einstein or Marie Curie — realistically recreated — explain relativity or radioactivity. But how do viewers react to these avatars? Does realism generate discomfort and thus mistrust, or can it actually improve understanding and the transmission of scientific information? A study published in the Journal of Science Communication (JCOM) explored how the realism of AI-generated avatars influences perceived trustworthiness in science communicators, also examining the role of avatars’ gender.

When West Virginia recently banned seven artificial food dyes in products to be sold within their borders, they joined an increasing number of individual U.S. states issuing their own regulations about food manufacturing practices, allowable ingredients, or product labeling. Consequently, food manufacturers must decide how to deal with different requirements in multiple markets. A new study from the University of Illinois Urbana-Champaign examines the various ways manufacturers respond to state regulations and what drives their choices.

A groundbreaking study from East China Normal University has leveraged AI to analyze classroom teaching patterns across China, revealing that teacher-centered instruction still dominates. This research utilizes a sophisticated AI-powered classroom intelligent analysis system to assess 1,008 recorded Chinese language classes spanning nine grade levels. Usage of AI enables effective systematic analysis of large volumes of classroom data, reducing time and human intervention.