Skip to content
aiqbench takeAI Products

aiqbench analysis

This article focuses on Calilia's "Nuwa" model, which is specifically designed to track the continuous spatial and temporal states of the human body rather than acting as a generic large language model. By integrating perception, modeling, decision-making, and execution into a closed loop for physical interactions, the system addresses the unique challenges of applying AI to embodied agents with irreversible consequences. The core innovation lies in treating biological tissue dynamics as computable data within a specialized multimodal framework distinct from standard digital human representations.
Source covering36Kr
What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

Calilia Closes the Loop on Physical AI: Turning the Human Body into a Computable Reality

36Kr1 min read

Wanmeijing AIMIRA constructs a 'Perception-Understanding-Decision-Execution-Feedback' product loop to address the core proposition of physical AI. As AI moves beyond screens, the rules change: real human bodies possess spatial structure, tissue states, and continuous temporal changes; while software errors can be regenerated, machine action errors may have irreversible consequences. The challenge for physical AI is not larger models but the entire chain of 'Perception-Modeling-Decision-Execution-Verification-Relearning.' Guangzhou Calilia Innovation Technology (Calilia International Group/Wanmeijing AIMIRA) has productized this chain: FAITH serves as the perception entry, the 'Nuwa' human real-world model handles modeling and understanding, medical-grade intelligent hardware and embodied devices execute actions, while effects and long-term changes feed back into the system. Dr. Jiang Yuzhen, founder of Calilia, stated: 'Nuwa is not a generic 'human large model' that answers everything, but a multimodal real-world model centered on the actual human body, incorporating both spatial structure and temporal dimensions. Digital humans solve 'how to present a person,' while we solve 'how to enable machines to continuously understand what changes are happening to this real person.'

Read the full article at 36Kr

Related coverage