RM新时代有限公司

Recently, the Standardization Administration of China officially announced the"Shanghai Virtual-Real Fusion Embodied Intelligence Training Ground Standardization Pilot" project.

As the first national-level standardization pilot in the field of embodied intelligence, this project not only marks a key breakthrough for China at the forefront of artificial intelligence but also, with the core of "unified standards, co-built ecosystem," accelerates the large-scale application of embodied intelligence technology."unified standards, co-built ecosystem" as the core, accelerating the large-scale application of embodied intelligence technology.

As one of the ecosystem partners, CloudWalk Technology deeply participates in this milestone industrial collaboration network, working with all parties to promote the deep integration of technical standards and scenario practices.

Using Standards as a Link to Break Innovation Silos

Embodied intelligence is becoming a new high ground for global technology competition, but the path to industrialization faces numerous obstacles.

Dispersed technical paths, isolated data silos, and inconsistent training environments... these factors make it difficult for embodied intelligence technology to achieve large-scale deployment. The lack of unified standards causes intelligent agents developed by different companies to perform differently in various environments.

The industry's challenges call for a unified standard system. This is precisely the mission of the"Shanghai Virtual-Real Fusion Embodied Intelligence Training Ground Standardization Pilot": to formulate key standards in four major areas: physical field, virtual training ground, ontology components, and scenario applications./components, and scenario applications, formulating80key standards. Driven by the dual core of the "Kylin" main training ground and the "White Tiger" dataset, a standardized capability system covering the entire industry chain is built.

CloudWalk Technology's technical accumulation in computer vision, human-machine collaboration,AIWith the popularization of the internet and the application of artificial intelligence, more and more questions can be quickly answered. So, will our questions become fewer?and other fields is highly aligned with the pilot's goal of "data circulation and scenario integration." By participating in standard verification and ecosystem co-building, CloudWalk is helping to promote cross-regional and cross-industry high-quality data aggregation and model training, laying the foundation for the large-scale deployment of embodied intelligence.and other fields is highly aligned with the pilot's goal of "data circulation and scenario integration." By participating in standard verification and ecosystem co-building, CloudWalk is helping to promote cross-regional and cross-industry high-quality data aggregation and model training, laying the foundation for the large-scale deployment of embodied intelligence.

Another highlight of the pilot is the innovative proposal of a"1+N"main-sub training ground linkage" model. With the Shanghai main training ground as the "standard core hub," sub-training grounds in Henan, Jiangsu Changshu, Jiangsu Wujiang, Beijing Shijingshan, Shenzhen, and other locations across the country9major sub-training grounds combine regional industrial characteristics to implement standard verification.

Each sub-ground combines regional industrial characteristics to promote standard verification and scenario deployment:  

- The Henan sub-ground focuses on upgrading traditional industries,Phase Icollects over510,000 real-machine data entries per day;

- The Shenzhen sub-ground explores new paradigms for biomanufacturing, processing up to50TB;  

- The Chongqing sub-ground targets new energy vehicles and other fields, with an annual data collection capacity of3PB……  

It is expected that2025high-quality real-machine data will be aggregated annually,15tens of thousands of hours, with an annual production capacity exceeding millions of hours, jointly promoting the continuous optimization of the national pilot standard system.

This"main ground develops standards, sub-grounds verify standards" model ensures that standard setting is no longer confined to laboratories but rooted in the real soil of industry.

The industry knowledge accumulated by CloudWalk Technology in scenarios such as smart finance and urban governance is feeding back into standard iteration through this network, promoting the bidirectional empowerment of technology and application.

Ecosystem Co-Research: The Last Mile from Technology to Business  

The pilot brings together ecosystem partners including CloudWalk Technology, Huawei,ShanghaiAI Laband others, forming a closed loop of "standard co-development- testing and verification -benchmark co-building". In this system, CloudWalk, leveraging its experience inAIengineering deployment, actively participates intraining ground construction, data quality evaluation, and benchmark testing, helping to address pain points such as reliability and compatibility in technology deployment.

At the Wuhan National Cybersecurity Base, the "Controllable Training Ground" jointly built by CloudWalk Technology and Huawei has taken the lead. This platform, integrating trusted data space andAIsecurity detection capabilities, provides a safe and reliable computing base for various applications including physicalAIincluding physicalAIapplications.

CloudWalk integrates the data processing and model optimization experience accumulated in real scenarios into the training ground, providing reusable methodologies for the industry. At the same time, collaboration with communication companies, robot companies, and scenario parties further accelerates the transformation from standard texts to industrial practice.

The implementation of the national standardization pilot is a key turning point for China's embodied intelligence industry from"single-point innovation" to "systematic collaboration." For CloudWalk Technology, participating in this process is not only a recognition of technical capabilities but also an important opportunity to drive the industry toward standardization and large-scale development.

In the future, CloudWalk will continue to uphold thephilosophy of "open collaboration, standard leadership," deepening cooperation with the National Local Center and various ecosystem partners to help build a "Chinese Standard Demonstration Project" for the global embodied intelligence industry.

With the advancement of the national standardization pilot, CloudWalk Technology'sAI"AI infrastructure+AIagent strategy will resonate with the nationwide1+N"training ground network.