Researchers at ETH Zurich are developing a digital co-pilot, or “assistant”, that helps specialists take a comprehensive approach when restoring historic sandstone buildings suffering from weathering damage.

The co-pilot combines artificial intelligence (AI) and extended reality (XR). An initial version was developed and implemented as part of a pilot project at Notre-Dame Cathedral in Lausanne.

The project offers an example of how digital tools can be used in built heritage conservation.

Imbuing light with "structure" adds complexity, mixing properties in exotic ways. In parallel, machine intelligence based on complex deep neural networks complexity for its learning, mimicking the complex connections of neurons in the brain. Now researchers have shown the exciting benefits at the interface of the two, from machines unravelling complexity in light, to optical systems enabling ultrafast learning machines. This symbiotic relationship is set to drive a new revolution in photonics.

Researchers at the University of Waterloo have developed a machine learning approach that could help public health officials anticipate disease outbreaks by analyzing vaccine skepticism on social media. Using the mathematical concept of tipping points, the team showed that online misinformation spreads much like a contagion and can signal heightened outbreak risk well before illness appears. Their model, tested on posts from California ahead of a 2014 measles outbreak, detected early warning patterns that traditional tweet-counting methods missed. This work supports Waterloo’s efforts to strengthen evidence-based decision-making and rebuild public trust in science, and the model could eventually be adapted across platforms like TikTok and Instagram to help identify communities nearing a critical shift in vaccination sentiment.

Touch is the sense that brings us into direct contact with reality, revealing shape, texture, and resistance. Designing soft sensors to mimic biological fingertips facilitates natural haptic communications in telerobotics and prostheses, but suffers from inaccurate tactile decoding. Scientists at Shanghai Jiao Tong University reported a 3D-architected pressure sensor featuring a hydrogel lattice encapsulated within an origami-inspired framework. The designed linear electro-mechanical responses enable sophisticated pressure input for robotic control and an intelligent fingertip for modulus detection.