This week’s list of top data news highlights from January 10, 2026 to January 16, 2026, includes disinfecting drinking water more safely by predicting harmful chemical byproducts and using underground road sensors to improve winter road safety and salt deployment.
1. Synching Robots Lip Movement with Audio
Researchers at Columbia University have developed a new AI-driven system that allows a humanoid robot, called Emo, to produce realistic facial expressions that move in sync with speech, making interactions feel more natural and human-like. Emo has a movable face covered in flexible synthetic skin, and its expressions are controlled by two AI models: one predicts how the robot’s mouth and face should move based on upcoming speech, while the other predicts how a human conversation partner is likely to react. Emo uses that prediction to adjust its own facial expressions in advance such as smiling back or softening its expressions
2. Filtering Wireless Phone Signals
Researchers at the University of Colorado Boulder, the University of Arizona, and Sandia National Laboratories have developed a new microchip that processes wireless signals by creating tiny, controlled vibrations on its surface. The chip generates these vibrations by converting electrical energy into mechanical motion using specialized materials layered onto a silicon base. These vibrations act as precise filters, separating useful signals from background noise, a function that currently requires multiple bulky components in smartphones. By integrating this capability into a single chip, the technology could enable thinner devices, lower power consumption, and faster wireless performance.
3. Learning Without Task-Specific Training
Researchers at 1X, a Norwegian robotics company, have developed a new AI system that lets their humanoid robot, NEO, plan its actions by predicting the outcome of its movements, rather than relying on task-specific training data. The system learns by watching large numbers of videos of people performing everyday activities, such as picking up objects or moving around a room. From those videos, it learns what usually happens next in a given situation and uses that expectation to decide how NEO should move. Because the robot is not trained on each task individually, it can handle new objects, environments, and everyday tasks it has never been explicitly taught.
4. Customizing 3D Objects Without Weakening Them
Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory have developed MechStyle, an AI system that helps people customize 3D-printed objects without making them fragile or unusable. When users change the look or shape of an object, the system automatically checks how those changes affect strength and function, adjusting the design to prevent breakage. Users can upload common items, such as hooks, glasses frames, or pillboxes, and modify their appearance using text or image prompts, while the system ensures the object can still be printed robustly.
5. Improving Winter Road Safety
The Maine Turnpike Authority has installed underground weather sensors beneath sections of the highway to improve winter road safety and make plowing and salting more precise. The sensors measure temperature at the surface and below the pavement, sending real-time data back to operations staff, who use it to decide when and where to deploy plow trucks and how much salt to apply. Officials say the system helps them respond more quickly to changing conditions, manage salt supplies more efficiently, and target problem spots before roads become dangerous, while still reminding drivers to slow down during winter weather.
6. Advancing Drone Sight in Harsh Weather
Chinese drone maker GDU has developed an AI-powered drone designed to operate in poor visibility, including fog, dust, and darkness. The drone uses optical sensors and AI image processing to improve what it can see in real-time, allowing operators to capture clear images day or night using standard, thermal, and night-vision cameras. It can also navigate using its onboard cameras when GPS signals are weak or blocked. The drone is designed to be used for infrastructure inspections, search-and-rescue missions, disaster response, and security monitoring in hazardous environments.
7. Helping Incarcerated People Prepare for Life After Release
A Los Angeles nonprofit has implemented a program in California prisons that uses virtual-reality headsets to give incarcerated people both a brief mental escape and practical preparation for life after release. The VR system immerses participants in realistic, everyday scenarios, such as job interviews, navigating public spaces, or reconnecting with family, offering hands-on learning that worksheets or lectures can’t provide. It also lets individuals visit calming outdoor environments, such as beaches in Thailand or streets in Paris, which organizers say has helped reduce disciplinary incidents.
8. Linking App Data to Google’s Gemini
Google has launched Gemini’s Personal Intelligence feature, which lets the AI assistant access apps like Gmail and Google Photos to provide more relevant, context-aware help. Unlike earlier versions that relied only on general web knowledge, it can reason across emails, photos, and search history to answer practical questions, such as finding a license plate in a photo or suggesting tires based on past trips. By combining context-aware reasoning with a user’s own data, Personal Intelligence makes Gemini more proactive, personalized, and useful for everyday tasks.
9. Upgrading Fighter Jets with AI
China has announced an AI system upgrade to its J‑20 stealth fighter jet to enhance performance in future air combat. The new system allows the jet to operate as part of a networked team of aircraft and drones that share data and coordinate actions in real-time, using AI to improve situational awareness and target planning. The AI enhancement boosts autonomous decision-making, helping the jet detect threats faster and maneuver more effectively. This shift toward AI-assisted operations reduces reliance on human pilots for split-second decisions and signals a move toward more automated, intelligent aerial warfare.
10. Discovering Hidden Toxins in Drinking Water
Researchers have developed an AI-driven system that helps water-treatment facilities disinfect water more safely by predicting when harmful chemical byproducts might form. The system uses sensors to monitor water quality, measuring factors such as chemical levels, pH, cloudiness, and disinfectant concentration, to analyze how chemicals will react with contaminants. It then recommends real-time adjustments to dosing. Unlike traditional fixed-dosing or trial-and-error methods, this adaptive approach kills pathogens while minimizing toxic byproducts.
