Anthropic 推 MHS 标准,让 Claude 操控实验室设备

Rohan Paul · @rohanpaul_ai · X·2026-08-28 03:53·9小时前
AI 导读

Anthropic 正测试让 Claude 控制机器人与实验设备的新方法,其 Model Hardware Standard(MHS)通过统一接口,让 AI 智能体操控实验室和工厂硬件。MHS 用通用驱动标准化命令、设备属性和安全限制,取代各仪器独立的编程接口,智能体可经 MCP、命令行或代码访问驱动。Anthropic 称早期项目已将集成工作从数周缩短至数小时或几分钟。

Rohan Paul@rohanpaul_ai
40AI 编辑部评分,满分 100

Anthropic 推 MHS 标准,让 Claude 操控实验室设备

2026-08-28 03:53· 9小时前
AI 导读

Anthropic 正测试让 Claude 控制机器人与实验设备的新方法,其 Model Hardware Standard(MHS)通过统一接口,让 AI 智能体操控实验室和工厂硬件。MHS 用通用驱动标准化命令、设备属性和安全限制,取代各仪器独立的编程接口,智能体可经 MCP、命令行或代码访问驱动。Anthropic 称早期项目已将集成工作从数周缩短至数小时或几分钟。

Anthropic is testing a new approach for Claude to control robots and lab equipment.

Its new Model Hardware Standard (MHS) will let AI agents control lab and factory hardware through 1 shared interface.

MHS is the missing interface between AI agents and physical equipment, using common drivers for commands, device properties, and safety limits.

Most instruments expose separate programming interfaces, so connecting a lab or factory can take weeks or months.

MHS standardizes that work around drivers exposing simple read/write commands plus reference files describing device capabilities, physical properties, and enforced safety limits.

Agents can access those drivers through MCP, command-line tools, or code, then sequence work across devices and adjust settings from live results.

Anthropic says early projects cut integration work to hours or minutes.

AnthropicToday, we're kicking off the first phase of the research preview for Model Hardware Standard (MHS): a new standard for AI agents to safely operate physical equi...