PulseAugur
LIVE 10:15:14
tool · [1 source] · · 中文(ZH) 灵巧手的三大路线之争:连杆、腱绳与直驱 — Chestnut
0
tool

Chestnut Robotics pursues hybrid dexterity for robot hands, blending tendon and AI

Chestnut Robotics is developing a dexterous robotic hand, opting for a hybrid architecture that primarily utilizes a tendon-driven system for fine manipulation. This approach aims to mimic human hand dexterity while incorporating AI for control. The company's founder, Evan Tao, a former key member of Tesla Optimus's dexterous hand team, highlighted the ongoing debate in robotic hand design between linkage, tendon, and direct-drive mechanisms. Chestnut's choice prioritizes the precision of tendon systems, addressing challenges associated with cost and control complexity in other designs. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT This hybrid robotic hand design, incorporating AI control, could advance the capabilities of humanoid robots in complex manipulation tasks.

RANK_REASON Product development announcement from a robotics company, not a frontier AI release.

Read on 36氪 (36Kr) →

COVERAGE [1]

  1. 36氪 (36Kr) TIER_1 中文(ZH) ·

    The Three Major Route Disputes of Dexterous Hands: Linkage, Tendon-String, and Direct Drive — Chestnut

    当下,有关灵巧手设计的核心迷思是:要不要像人手?围绕这一命题,诞生了连杆、腱绳、直驱三种设计方案。 其中,“连杆”最不像人手,但胜在成本低、易于控制;“腱绳”最像人手,可以做精细化操作,但成本高、控制难。“直驱”则是一种折中方案,将驱动单元直接集成在每个关节上,但成本不低,同时力传导效率和热管理上仍然面临工程层面的挑战。 混合架构路线,则是近期兴起的灵巧手技术解决方案。 Chestnut Robotics创始人、前Tesla Optimus灵巧手核心成员Evan Tao介绍,当下团队已经选择了混合架构路线,以可以完成精细化操作的腱绳结构为主,辅以AI控制