This week’s list of top data news highlights from January 3, 2026 to January 9, 2026, includes developments in cooling data centers and Lego’s new Smart Brick, a motion‑sensing component that adds real‑time interactivity to traditional builds.
1. Detecting Cancer with AI Generated Sensors
Researchers at MIT and Microsoft have built an AI system called CleaveNet to make hidden signs of disease easier to detect and target. Many diseases, including cancer, change the body’s chemistry in subtle ways by increasing the activity of certain enzymes, but those changes are difficult to observe directly. CleaveNet designs short protein sequences that are cut only when such enzymes are active, effectively turning enzyme activity into a detectable signal. This makes otherwise invisible disease processes measurable and allows drugs or imaging agents to remain inactive until they encounter the biological conditions they are meant to respond to, improving precision and reducing unintended effects.
2. Automating Construction Vehicles
American construction company Caterpillar recently unveiled its new fleet of fully autonomous construction machines that integrate machine learning, computer vision, and edge computing. Equipment such as excavators, haul trucks, and dozers equipped with LiDAR, radar, GPS, and high‑resolution cameras generate a continuous 360‑degree digital model of the jobsite, allowing them to carry out autonomous trenching, grading, loading, and material transport.
Finnish eyewear company IXI has developed autofocus glasses that use eye‑tracking sensors to capture exactly where the wearer is looking. The system feeds gaze coordinates into a controller that reshapes the glasses’ liquid‑crystal lenses in real-time to match the needed prescription. Instead of relying on fixed optical zones, such as bifocals, IXI’s glasses dynamically adjust the entire lens surface to keep images sharp. Onboard electronics process eye‑movement data hundreds of times per second, allowing instant focus shifts between near and far objects.
4. Finding the Perfect Water Droplet Size to Cool Data Centers
Researchers at Virginia Tech have built an AI system inspired by Google’s DeepMind’s AlphaGo, to explore how different water droplet shapes affect data center cooling efficiency. The team trained machine‑learning models on droplet‑size datasets collected from 25 prior studies to understand how different liquids form droplets and transfer heat. Their model identifies optimal droplet characteristics for high‑efficiency spray cooling, a key technique for more efficient data‑center heat management.
5. Designing Software to Identify Aircraft Weaknesses
Researchers at the China Aerodynamics Research and Development Centre have built a simulation tool that makes it faster to analyze aircraft designs. The system, called PADJ‑X, analyzes data about an aircraft’s airflow, structural strength, and radar visibility, then applies adjoint optimization—a method that shows how changing one part of a design affects the rest without rerunning the full simulation—to calculate how small adjustments across hundreds of design variables influence flight performance.
6. Picking Apples with a Robotic Hand
Researchers at Washington State University have developed an inflatable robotic arm that apple‑picking robots could use in the future to gently harvest fruit. Mounted on a metal base, the arm uses computer vision to locate apples, with an onboard camera feeding visual data to an AI model that determines their precise position and orientation. Controlled inflation and deflation let the arm extend, grip, and retract without harming the fruit or branches. The team is also integrating it with a Cornell‑built movement platform to support fully autonomous harvesting.
7. Training Humanoid Robots for Factory Work
Boston Dynamics has built an AI‑powered humanoid robot that learns factory tasks by watching humans. The robot processes video and sensor data to map human actions into a database of reusable skills. Its AI model recombines those skills to perform new tasks without manual programming. Onboard cameras and depth sensors maintain a live 3D workspace model, allowing real‑time adjustments in the robot’s grip, balance, and motion.
LEGO recently introduced the Smart Brick, a standard 2×4 block equipped with sensors and a custom chip that lets Lego structures respond directly to play. Each block houses motion and light sensors, LEDs, a small speaker, and wireless power, and communicates with specially tagged Minifigures and tiles over a short-range Bluetooth-based network called BrickNet. By processing movement and proximity on the blocks themselves, the structures can trigger sounds and lighting in real-time without an app, hub, or Internet connection—keeping play hands-on while making physical builds more expressive, replayable, and story-driven.
9. Building More Efficient Super Computers
Scientists at Sandia National Laboratories have built an algorithm that allows neuromorphic hardware—computers designed to work like networks of brain cells—to solve complex equations used to model how fluids move. The system converts these equations into patterns of spiking-neuron activity and processes them through low-power, parallel circuits that mimic simplified brain networks, allowing it to deliver accurate results while using far less energy than traditional CPU‑ or GPU‑based supercomputers.
10. Communicating Pill Capsules
MIT engineers have designed a pill that transmits a radio‑frequency (RF) signal from inside the stomach to confirm medication ingestion. A biodegradable antenna remains inactive until stomach fluid dissolves the pill’s coating, triggering a brief RF transmission to an external receiver. The material fully breaks down in the stomach, while a tiny RF chip passes safely through the digestive tract to collect additional data on the patient’s health. This low‑cost system provides precise adherence data for patients requiring strict dosing and integrates easily into existing pill‑manufacturing processes.
