Neura Robotics 与慕尼黑工业大学建机器人训练场，约 100 台机器练习物理技能
This is a gym for robots 🦾👀 Neura Robotics and the Technical University of Munich created a facility where around 100 machines practice physical skills, from handling objects to using tools. Human trainers guide their movements while sensors capture data about motion, contact, and force. That experience helps train AI to connect what robots sense with what they should do next. Simulations expand the training, and Neura says learned skills can be shared with other connected robots. The goal is to make robots reliable enough for everyday work, where repeating a task consistently matters far more than getting it right once.
Neura Robotics 与慕尼黑工业大学建了一处机器人训练场，约 100 台机器在其中练习抓取物体、使用工具等物理技能。人类训练师远程引导动作，传感器记录运动、接触和力数据，用于训练 AI 把机器人感知与下一步动作对应起来；仿真补充变化，Neura 称学到的技能可通过其云平台共享给其他联网机器人。
原文展示了约 100 台机器人集中训练的场景和传感器采集动作、接触、力的细节，适合做热点评论，聊机器人技能如何通过云端共享给其他机器。
粉丝 8.6万
The latest in AI and robotics, explained Founder & CEO of @The Rundown AI Zurich 🇨🇭 / NYC 🇺🇸
- 浏览
- 980
- 点赞
- 57
- 评论
- 1
- 分享
- 0
- 收藏
- 8
数字来自平台，为最近一次抓取时的数据。
以下是视频口播的自动转写，可能有错字。
So, this is a gym for robots. Inside, around 100 machines are practicing things we do without thinking, like picking things up and putting them down.
But each exercise also generates training data that ai needs to operate the physical world. Neuro robotics and the technical university of munich built this facility so robots can practice skills that humans take for granted.
Tightening a bolt, for example, requires more than recognizing where it goes. A robot has to align the tool, apply the force, and control its movement throughout the whole time.
There are human trainers that help by remotely guiding robots through tasks, while sensors record movements, contact, and force.
And those demonstrations help train ai models to connect what a robot senses with what it should do next. Computer simulations add variations, which researchers then test on real machines.
And the experience doesn't have to stay with one robot. Neuro says its cloud platform makes learned skills available to other connected robots, so entire fleets could benefit from the exact same training.
The challenge is making those skills reliable enough for actual work, because doing correctly thousands of times is what makes a robot actually useful.
If you want to stay in the loop with the latest in AI and robotics, follow for more.
来源：TikTok · Rowan Cheung · tiktok.com