Back to timeline

Mon, July 2017:12ToolsInfra & costRobotics & embodiedInfra & cost guide

Qingyi Innovation Launches Dino OS for 80cm Plush Humanoid Robot Amoo

Decision Brief

What changedQingyi Innovation unveils self-developed robot OS Dino OS to solve stability issues in consumer robot iteration.
Why it mattersDino OS's three-layer architecture and dual-brain closed-loop design let developers reuse behavior units, slashing development cycles from years to months.
Who should careAI coding tool users
Affected stackNo specific stack identified
Source confidenceMedium · Reliable media or first-hand reporting

On July 14, Qingyi Innovation released Dino OS, built on a three-layer architecture: Infra, Brain, and Module. The Infra layer uses zero-copy data serialization and shared memory communication for parallel data flow in perception, decision, and execution, with built-in compute isolation for critical tasks like obstacle avoidance. The Brain layer features a 'dual-brain closed-loop' design: a slow brain for scene and task understanding, a fast brain for instant social responses, and a cerebellum for motion control and force feedback, ensuring millisecond-level execution. The Module layer provides airtight compilation environments, packaging complex perception-to-execution pipelines into reusable behavior units with unified interfaces. The first product running Dino OS is Amoo, an 80cm-tall, bipedal, plush-covered robot. Previously, consumer robots lacked a unified OS, causing stability issues when adding features (e.g., falling after dancing). Dino OS ensures incremental development without breaking stability via fixed communication channels, consistent control, and interrupt mechanisms. This cross-body portability can compress development cycles from years to months, freeing developers to focus on emotional and interactive experiences.

Summary basis: full article readCompiled from the source scope noted above; the original remains authoritative.

Sources

  • 36氪

    Chinese tech/startup coverage; AI signal filtered from broader business news.

  • 36氪

Related intel

留言

登入后即可留言,和其他 builder 交换实测心得。

还没有留言,抢头香。