This Robot Carries a Human

Unitree Technology’s B2-W robot dog is proving its mettle—literally. A 12 months post-mass production, this machine now flips, slides, and scales steep slopes with ease, all while carrying a 40-kilogram load and jumping from 2.8 meters in height. It’s not only a flashy demo; the robot dog’s real value lies in its potential for industrial applications, from hazardous inspections to fireplace rescues.

This mix of agility, power, and endurance is quickly turning the B2-W right into a tool able to handling the hardest physical challenges. In other words, the long run of robotics just got an entire lot more mobile and functional. Read more.

Researchers at Shanghai Jiao Tong University and Alibaba unveiled MatchVision, an AI system trained on SoccerReplay-1988—the most important dataset of its kind, featuring 2,000 matches and over 3,300 hours of European league footage. MatchVision identifies 24 game events, including fouls, and evaluates their severity with as much as 84% accuracy.

This AI system not only tracks on-field actions but additionally generates human-like commentary by analysing context and technical play. Outperforming existing models, MatchVision could streamline match evaluation, produce automated highlights, and assist referees. The dataset and model are set to debut on GitHub, promising broader access to advanced sports analytics. Read more.

Nvidia’s Jim Fan projects that nearly all of embodied agents will first take shape in virtual environments, before being zero-shot deployed in the actual world. These agents, linked by a “hive mind,” will exchange latent embeddings to synchronize actions in multi-agent tasks. A glimpse into this future is the City of Tokyo’s 3D digital twin, offering a high-res simulation of the whole city for download—only one example of how physical spaces are increasingly migrating into the digital realm.

On this landscape, robots will train not in isolation but inside an enormous “iron fleet” running on real-time graphics engines, creating trillions of coaching tokens at scale. Nvidia’s Santa Clara headquarters was designed and rendered in Omniverse, its GPU-accelerated platform, before any physical construction began. Read more.