Reflex Robotics

United States

公司簡介

Reflex Robotics is a US company building wheeled humanoids for industrial work, founded by engineers from Anduril, Oculus, Amazon, Waymo, Skydio, Boston Dynamics, ASML, Jane Street and Joby Aviation, and backed by Khosla Ventures, Liquid 2 Ventures, Collab Fund, 8090 Industries,…

機器人媒體與亮點

精選機器人影片

暫無精選機器人影片

🎬

暫無影片

該公司尚未關聯精選機器人影片。

暫無影片

機器人概覽

公司:Reflex Robotics

公司狀態:N/A

成立年份:N/A

融資階段:N/A

員工數:N/A

主要機器人方向:

暫無主要機器人方向資訊。

國家:United States

總部:N/A

官網:造訪官網

相關新聞

2026年10月20日3w ago

Germany Signs Industrial AI Pact With NVIDIA, Backs Agile Robots

Germany signed a Joint Declaration of Intent with NVIDIA on physical AI, as NVIDIA committed nearly US$10 billion across four German startups including humanoid-robot maker Agile Robots, and Siemens showcased an autonomous factory cutting engineering effort by up to 85 percent.

4分鐘閱讀

2026年7月27日

Robots Can See. This Shanghai Lab Just Published 30,000 Hours of Teaching Them to Feel

Shanghai startup NeoteAI (Xinzhi Embodied Intelligence) and Fudan University's Institute of Trustworthy Embodied AI released three technical reports on July 26 built on what they describe as more than 30,000 hours of visual-tactile interaction data, around 1.4 million manipulation clips across 450 real long-horizon tasks. The stack spans a tactile dataset (NeoData), a unified tactile representation model (NeoForce), an anticipatory visuo-tactile-language-action model (VTLA) reporting 85 to 99 percent success on contact-rich tasks, and a tactile world model (TWAM), with data and models pledged as open source. It is the most aggressive public claim yet that touch, not vision, is the next scaling frontier in robot learning.

4分鐘閱讀

2026年9月2日

Humanoid Robot Soccer Just Became the Industry's Most Honest Benchmark

A Tsinghua-led team published a reinforcement learning controller in Science Robotics that fuses vision and motion on Booster T1 humanoids, cutting ball-tracking error 46 percent and reaching a 90 percent kick rate with zero real-world fine-tuning. The platform math underneath matters as much: Booster hardware now carries most of competitive robot soccer, and the training method points at every dynamic task humanoids will face.

4分鐘閱讀