At the offices of Chinese startup DexForce, a screen shows a ​virtual robot placing objects on a table. It repeats the task as the angle of its arms, the ‌colour of the table and the objects themselves change.

At the offices of Chinese startup DexForce, a screen shows a ​virtual robot placing objects on a table. It repeats the task as the angle of its arms, the ‌colour of the table and the objects themselve...

By Laurie Chen SHENZHEN, Oct 15 (Reuters) - At the offices of Chinese startup DexForce, a screen shows a virtual robot placing objects on a table.

ロイター通信は、DexForceが仮想データを使ってヒューマノイドロボットを訓練し、コスト削減を図っていると報じた。 記事で示されたのは、実世界のデータ収集や実機訓練の負担を抑える狙いである。 一方で、訓練手法の具体条件や効果検証の詳細は限定的である。

At the Shenzhen offices of Chinese startup DexForce, a virtual robot repeatedly places objects on a table as the angle of its arms, the table’s colour and the objects change. The company is betting that simulated training can help solve a major challenge for the humanoid-robot industry: teaching machines to perform a wide range of real-world tasks without human assistance. This is reported by Reuters. China has developed humanoids that can outperform people in running and jumping. The harder commercial test is whether they can reliably do useful work in factories, offices and homes. Many Chinese robot makers are trying to meet that challenge by collecting large volumes of data from workers repeatedly performing tasks in real-world settings. DexForce instead relies on synthetic data generated in virtual environments to reduce costs and improve training efficiency. Chief Executive Jia Kui told Reuters the approach could accelerate a breakthrough in robot capabilities, akin to the one ChatGPT triggered after its 2022 release helped propel generative AI into the mainstream. “Trying to achieve ‘physical artificial general intelligence’ purely through real-world data collection would require 10 trillion samples … manual (data) collection for infinite human tasks is fundamentally inefficient,” – Jia Kui Jia was...

跨维智能（DexForce）は2026年10月12日、ロボット向けの3D世界モデル「UNITAS」を発表した。 leaderobot.comによれば、同社はこれを「全球首个機器人任务原生3D世界模型」と位置づけ、Physical AI向けの統一表現空間を探るとしている。 原典は二次報道であり、モデルの実装詳細や評価条件は限定的である。

At Beijing's World Robot Conference, Chinese humanoid robot manufacturers demonstrated a shift toward commercial deployment on assembly lines and in warehouses, backed by massive investor interest and...

Extracting Force-informed Actions from Kinesthetic Demonstrations for Dexterous Manipulation

翻訳：コンビニエンスストアにDEXFORCE W1が展示されています。セブンイレブンにこれを置きたいですか?