INDEPENDENT MARKET INTELLIGENCE / CHINA ROBOTICS
industrial

Astribot details CoMind framework for multi-robot collaboration

QbitAI reports that Astribot’s CoMind demonstrations improved selected task results while reducing the amount of information transmitted.

Software & embodied AIRobot makersHumanoids & embodied AI
ROBOPALLAS · 2026-11-16
Two humanoid robots coordinate a pen-pickup task across a visual barrier by sharing a short signal; conceptual illustration, not Astribot hardware.
AI-generated schematic illustration.

Astribot says its CoMind (Be My Eye) framework is designed for multi-robot coordination, letting robots exchange task-relevant information while making their own action decisions, according to QbitAI.

In a demonstration described by QbitAI, two robots on opposite sides of a visual barrier coordinated a pen-pickup task. The report says task success increased from 0% to 100% when they shared key information.

In a separate box-placement task, QbitAI reports accuracy rising from 38.9% to 88.9%, while the amount of information transmitted fell by 80%. The figures describe demonstrations in the report rather than a broad, independently validated deployment benchmark.

Companies in this report

Supplier profiles in the RoboPallas directory

Source: QbitAI ↗

← More reports