Independent global robotics news from Article module 102.
Each record is classified by both destination market and product layer, including complete robots and core components.
Running the Vision-Language-Action VLA model on RTX 4090 workstations is no longer a particularly difficult task. But the real difficulty is: how to get hundreds of milli...
Just a few weeks before the company's latest round of negotiations with the U.S. government, and just a few days before a security incident that has sparked industry-wide...
For a long time, mobile AI services have been mostly a "cloud story": mobile phones collect voice, images, text, location and context, send requests to servers, such as c...
As a result, the entire network stayed silent all day except for holding the FIFA prize. Just recently, sources said: "Don't wait, Opus 5 is delayed". The reason was that...
Hole holes are common and important characteristic information in metal parts and play a decisive role in the production and assembly of products. The method of reconstru...
Traditional 3D Gaussian reconstruction follows a fixed paradigm: depending on multiple images, SFM camera identification, multiple viewflow constraint generation optimiza...
Stripe is reportedly in talks to acquire OpenRouter. According to The Wall Street Journal and The Information, the price of the transaction is likely to be close to $10 b...
A few days ago, one of the most crowded concepts at WAIC 2026 was that Agent Computer chips began to take off from the rack and become a real engineering issue. Getting a...
Over the past two years, if one of the most crowded racetracks at the robotics fair were to be chosen, it's probably the gymnastics, running, dancing, back-and-forth, fro...
The AutoMIA system, developed by the team led by Liu Xiaobo, can convert any two images into 3D artwork with mirror illusion effects, achieve visual representation of opp...
The question is, is there a third way besides the Scaling Parameters and the Scaling Chain? The WeLM team gave a whole new answer to the problem of folding the extra forc...
In front of the white background, four new Coffield Prize winners lined up and sat down: Liu, John Pardon, Jacob Tsimerman, and Rainbow. Among them, Liu and Rainbow becam...
Editor's note: In June 2026, GLM-5.2 was released. This large-scale MoE model performed exceptionally well, shocking even the Next.js author, Vercel founder Guillermo Rau...
On the other hand, Claude, who can only run Haiku upgrades to the Opus 4.8 and Sonnet 5 series, can not only play Gmail, calendar, Slack in conversations, but also expand...
On July 15, former OpenAI CTO Mira Murati released the first model of Thinking Machines Lab, Incling, in San Francisco. With $2 billion in funding, the 975B parameter, 10...
A Skill will work normally this week, not next week. The search interface will be restricted, the API will be updated, and the page structure may change abruptly. The dev...
It addresses the problem of computational efficiency and residual occupancy in multi-modal large model training that is difficult to reconcile, with the question of relia...
Just today, GPT-5.6 Pro falsified the thirty-year conjecture of Dinitz-Garg-Goemans in the field of graph theory. His evidence is extremely concise: a graph, the cost of ...
On July 23, justin, the co-founder of AI startup Capy.ai, posted a post on X, and the team simply threw a post about the "randomest breakthrough in mathematics history" i...
A product introduction 3d structural light is mainly aimed at the three-dimensional reconstruction of dimensional reflective silicon objects, commonly used in the field o...