This week’s list of top data news highlights covers May 24, 2025 to May 30, 2025 and includes articles on using AI to track food waste and streamlining hospital operations with help of robots.
1. Generating Business Avatars
Akool, a California-based software company, has launched an AI-powered video tool that creates realistic digital avatars for sharing company-wide messages and training content across global teams. Called Live Camera, the tool turns a written script into a human-like video, delivered by an avatar that speaks naturally, gestures, and matches lip movements in multiple languages. Unlike live video calls or pre-recorded presentations, it doesn’t require the speaker to appear on camera or record separate versions. Instead, it generates localized videos automatically, helping companies save time, cut translation costs, and deliver consistent, engaging content across time zones.
2. Building Efficient AI Chips
Cerebras, a semiconductor company based in California, has developed the world’s largest computer chip, designed specifically to speed up AI tasks. Its size allows all the processing and memory to happen on a single chip, avoiding the delays that come from splitting work across multiple chips. This design recently set a new speed record for running Llama 4, processing 2,500 tokens per second—more than twice as fast as NVIDIA’s top chip. It could be especially useful as AI systems take on more complex, multi-step tasks.
3. Streamlining Hospital Operations
Diligent Robotics, an AI robotics company based in Texas, has deployed service robots in U.S. hospitals to help address workforce shortages by automating routine tasks like delivering medications and supplies. Its robot, Moxi, uses computer vision, obstacle detection, and autonomous navigation to move through crowded hospital corridors, press elevator buttons, and open doors. Recent improvements in AI planning and perception allow Moxi to operate in older hospitals with inconsistent layouts and limited automation infrastructure. Hospitals using these AI-powered service robots report operational efficiencies and time savings as high as 40 percent.
AdAstra Sustainability, a Swiss climate data company, has launched Orbae, an open-data platform that shows where forests are being cleared or land is being converted for farming. Using satellite images and models, Orbae tracks 11 major crops—including soy, palm, and sugarcane—across more than 30 countries. Companies can use Orbae to see the environmental impact of their supply chains and include land-use emissions in their carbon reporting, in line with international climate standards.
The City of Toluca, Mexico, has partnered with India-based neuralapps.ai to launch an AI-powered platform that monitors and manages water use across the city. Designed for public utilities, the system uses real-time data and AI algorithms to detect leaks, flag abnormal consumption, and predict maintenance needs, reducing water loss and service disruptions. It also generates automated, audit-ready compliance reports, helping Toluca improve transparency and efficiency.
Becton Dickinson, a medical technology company based in New Jersey, has launched an AI-powered monitor designed to detect early signs of blood pressure drops during high-risk surgeries. The device uses noninvasive sensors and machine learning to track real-time patient data and spot signs of instability before symptoms appear. In early trials, it predicted these drops with over 30 percent greater accuracy than standard monitors, allowing surgeons to respond earlier and adjust anesthesia more precisely.
Rendever, a virtual reality company based in Massachusetts, has introduced an immersive platform that enables aging veterans to revisit meaningful places and experiences through shared VR sessions. Using motion-tracked headsets and a curated library of 360-degree environments, the system enables group-based virtual travel—from hiking the Grand Canyon to visiting national memorials—within senior care facilities. At the Veterans Community Living Center in Colorado, the platform is helping residents combat social isolation, stimulate memory recall, and process emotional milestones through guided sessions.
8. Improving Accessibility for Blind Farmers
Girish Badragond, an inventor based in Karnataka, India, has developed a smart farming stick that uses soil sensors and weather data to support blind and partially-sighted farmers. The device translates environmental conditions into audio messages and vibrations, enabling users to assess crop health and make timely decisions without visual input.
9. Understanding Drought Traits
Researchers from Hebrew University and the Volcani Institute in Israel have developed a drone-based system that uses AI to identify drought-resistant wheat. The drone captures images of wheat fields and, using machine learning, estimates traits like chlorophyll levels and water use across 300 different types of wheat. This method improved water-use estimates by 28 percent compared to traditional manual testing. It also helped scientists find 16 specific parts of the wheat genome linked to drought resilience.
Zest, a UK climate-tech start-up, has built an AI system that watches food-factory production lines and spots edible products that would normally be thrown out, such as broken chocolates or items with too little shelf-life left for retailers. Cameras feed images to machine-learning software that flags this hidden waste in real-time, so staff can redirect it for quick redistribution instead of disposal. In a two-week pilot at a Nestlé plant the tool cut edible waste by 87 percent, a saving that could reach 700 tonnes of food a year (about 1.5 million meals) and avoid 1,400 tonnes of CO₂ emission.
