This week’s list of top data news highlights covers April 12, 2025 to April 19, 2025 and includes articles on screening babies for cataracts with AI and wearable devices that are making navigation easier for visually impaired users.
Dartmouth researchers have developed Therabot, an AI chatbot designed to deliver evidence-based mental health support and help address a shortage of human providers. To train it, the team created a custom dataset of therapeutic scenarios to ensure the model could respond in line with clinical best practices. In the first clinical trial of its kind, users with anxiety, depression, or eating disorders who engaged with Therabot over eight weeks saw meaningful improvements in symptoms, including a 51 percent reduction in depression scores.
2. Enhancing Navigation for the Visually Impaired
Researchers from Shanghai Jiao Tong University and the Hong Kong University of Science and Technology have developed a wearable system that helps blind and visually impaired users navigate using AI-powered glasses, audio cues, and vibration feedback. A camera captures real-time surroundings, and machine learning detects objects like walls and doors, guiding users with audio beeps every 250 milliseconds. Additional wrist-worn sensors vibrate as users approach obstacles or reach for objects. In trials, participants navigated indoor spaces 25 percent more efficiently than with a cane.
3. Deploying Sensors for Climate Resilience
The Brazilian city of Pindamonhangaba is using AI-linked sensors to monitor floods, detect wildfires, and track mosquito outbreaks as part of its climate adaptation strategy. After a major flood in 2024, the city installed water-level sensors in high-risk areas to trigger early alerts, helping authorities avoid over 1 million reais (approximately 200,000 U.S. dollars) in damage and emergency costs. Built by local startup iNeeds, the same platform is now being expanded to include wildfire detection in rural zones and mosquito tracking to prevent dengue outbreaks.
A nonprofit platform called Love Lost is using image recognition to help reunite lost pets with their owners. The system identifies animals by analyzing features like eye shape, markings, and tail curvature. Since launching in 2021, it has reconnected 100,000 pets with their families and is now used by over 3,000 shelters nationwide.
5. Transferring Data at Lightning Speed
TDK, a Japanese electronics company best known for being an Apple supplier, has developed an optical technology that could transfer data ten times faster than current electronics. The core of the advance is a spin photo detector—a device that receives light-based signals and converts them into electrical signals so chips can process them. In systems that use optical interconnects to move data between processors, this kind of detector plays a critical role at the receiving end. TDK’s version is unique because it combines optical, electronic, and magnetic materials to enable ultrafast response times of 20 picoseconds, or 20 trillionths of a second, while using less power than conventional designs.
6. Integrating AI into the Classroom
Ten New Jersey school districts are using state grants to integrate AI tools into teaching, applying them to generate quizzes, tailor reading samples, and support student writing. At Gateway Regional High School, one of the grant recipients, 65 percent of teachers have already incorporated AI into their lessons. In one example, students are using AI chatbots in French class to practice conversation skills, while teachers use the tools to personalize instruction and spend more time supporting individual students.
7. Screening Newborns for Cataracts
UK hospitals are testing an AI-powered screening tool to detect cataracts in newborns, the leading cause of preventable childhood blindness. Developed by an eye surgeon at a teaching hospital in Cambridge, the system is built into a handheld device that captures images of infants’ eyes, which are then analyzed by AI models. Trained on more than 46,000 images, the tool is currently being trialed in maternity wards across the UK as part of a national study expected to screen over 140,000 babies.
8. Tracking Homelessness with Better Data
Multnomah County, Oregon, has launched a new public dashboard to provide more consistent and detailed data on homelessness in the county. Updated monthly, the tool draws from expanded outreach efforts and a wider network of service providers to capture individuals who are unsheltered, in shelters, or whose housing status is unknown. In February, it recorded nearly 15,000 people experiencing homelessness, an increase driven by both rising housing instability and improved data collection. County officials say the dashboard offers a clearer baseline to guide policy decisions and increase public accountability.
9. Editing Motion Capture with AI
Researchers at Peking University have developed an AI tool that edits 3D motion data, which captures how people move using motion sensors and is used to animate characters in games, VR, and training videos. Normally, editing these movements requires either re-recording the motion or manually adjusting it in animation software. The new system, called MotionCutMix, lets users make changes with simple text prompts like “raise left arm” or “walk slower.” It blends parts of different movements and smooths transitions automatically, making the process faster and more flexible.
10. Revealing Earth’s Hidden Structures from Space
NASA is preparing to launch the first quantum gravity sensor in space, designed to detect tiny shifts in Earth’s gravity by tracking the motion of atoms inside the device. By measuring how their movement changes as the satellite passes over different regions, the sensor can detect variations in subsurface mass, revealing features like underground water, minerals, or tectonic activity. Smaller and more sensitive than current systems, the technology could reshape how we monitor Earth’s hidden structures and manage natural resources.
