Facing the large-scale implementation of industrial embodied intelligence, GroupCore Technology has launched the cloud-native industrial AI twin platform, SpatialTwin.
Mobile robots flexibly shuttle between equipment to carry material boxes, humanoid robots sort products beside the conveyor belt, and robotic arms wave at high speed 24 hours a day... The industry generally expects that industrial scenarios will be one of the first areas where embodied intelligence can be applied. However, whether the existing industrial system can enable these intelligent agents to seamlessly integrate and collaborate efficiently has become a key pain point in the industry's implementation process.
Based on this, on November 6th, GroupCore Technology officially launched SpatialTwin, a cloud - native industrial AI twin platform specifically designed for industrial embodied intelligence, at the 2025 World Internet Conference. Meanwhile, SpatialTwin was selected as one of the top ten debut achievements at the "Internet Light" Expo of the 2025 World Internet Conference.
GroupCore Technology's SpatialTwin was selected as one of the top ten debut achievements at the "Internet Light" Expo of the 2025 World Internet Conference. Photographer: Huang Hang
Cloud - native technology brings new possibilities for human - machine collaboration in factories
It is understood that SpatialTwin is a product under GroupCore's SpatialVerse platform. It is the world's first cloud - native digital twin platform capable of real - time simulating the dynamics of real industrial environments and supporting the large - scale operation of intelligent agents. Relying on GroupCore's self - developed spatial large - model and integrating technologies such as digital twin, AI, and real - time rendering, SpatialTwin has built a full - link service system covering factory design and planning, simulation verification, smart operation and maintenance, and embodied robot simulation training. It is committed to reconstructing the human - machine collaboration model in industrial manufacturing scenarios.
"With the current rapid development of embodied intelligence, the industrial scenarios adapted to it must break the traditional layout centered on 'humans' and 'traditional equipment'. The key innovation of SpatialTwin lies in its cloud - native architecture," said Wu Kailiang, the senior vice - president of GroupCore Technology and the product leader of SpatialTwin. Most industrial twin platforms on the market are still mainly deployed locally, with problems such as complex deployment, limited computing power, and poor information sharing. SpatialTwin, based on cloud - native technology, achieves "zero - deployment". It can not only call high - density simulation computing power on demand to meet the computing power requirements of massive intelligent agents in the perception and reasoning process but also fully leverage the operation and calling efficiency of large AI models. At the same time, it supports all participants to seamlessly share information and collaborate in real - time, greatly improving the operational efficiency of industrial scenarios and reducing costs.
Caption: Wu Kailiang, senior vice - president of GroupCore Technology, shares the five key aspects of SpatialTwin. (Photographer: Huang Hang)
Facing the complex operating states of tens of thousands of devices and robots operating autonomously and collaboratively in future smart factories, SpatialTwin has focused on systematically reconstructing the industrial space for human - machine collaboration since its design. The platform deeply integrates GroupCore's spatial large - model capabilities, supports the access of multi - source heterogeneous data, and can quickly and automatically reconstruct three - dimensional scenes, creating a more adaptable industrial space foundation for human - machine collaboration in the era of embodied intelligence.
On this basis, SpatialTwin integrates digital twin technology. Through real - time rendering and physical simulation technology, it can accurately simulate the lighting, materials, and physical properties in industrial scenarios, and replicate real industrial scenarios, equipment, and processes in the digital space at a 1:1 ratio, truly building a dynamic and interactive digital twin factory.
Different from previous digital twin technologies that mainly focused on the dynamic monitoring and analysis of production factors, with SpatialTwin, factory operators can not only grasp the operating states of equipment and robots in real - time and accurately predict production trends but also let robots simulate real - world operation processes in the digital space and complete the full - process pre - rehearsal of "perception - understanding - decision - action", greatly reducing the trial - and - error costs in the actual production of robots.
Finally, just as humans need training before taking up a job, embodied robots also need professional "learning and training" before "going to work in the factory". SpatialTwin combines the synthetic data training solution of SpatialVerse. Through data generalization and scene randomization technology, it simulates the complex environments and diverse interaction tasks that robots may encounter during operation, creating a professional simulation training ground for industrial robots. The two build a closed - loop system of "reality - virtuality - reality", promoting the all - round evolution of robots in the fields of general intelligence, special intelligence, and multi - robot collaborative intelligence.
Caption: The digital twin screen of GroupCore Technology's SpatialTwin
Strategic cooperation with Hangcha Group and Standard Robots
At the debut event of SpatialTwin, GroupCore Technology also officially signed strategic cooperation agreements with enterprises such as Hangcha Group and Standard Robots to deepen cooperation in the field of industrial embodied intelligence.
As an important provider of intelligent logistics solutions, Hangcha Group is fully committed to building a full - scenario intelligent logistics service system. Currently, SpatialTwin has provided it with intelligent planning and 3D visualization design services for logistics warehouses, reducing the factory design cycle of regular project plans from one week to two hours, with a significant improvement in efficiency. According to the cooperation plan, the two parties will conduct in - depth cooperation around the integrated intelligent logistics solution of "planning, implementation, and operation and maintenance" in the future, and focus on promoting the intelligent upgrade of logistics humanoid robots to help the full implementation of intelligent logistics scenarios.
Tao Yikun, the person - in - charge of Hangcha Group's embodied intelligence project, said that Hangcha Group is accelerating its transformation from an overall intelligent logistics solution provider to a digital twin logistics solution provider. GroupCore Technology's mature solutions in the field of industrial AI twin will provide key support for the two parties to layout the large - scale and autonomous application of embodied intelligence, accelerating the intelligent transformation of the logistics industry.
Standard Robots is a global leading provider of industrial intelligent robot solutions, and it highly synergizes with SpatialTwin in terms of technical architecture and application scenarios. Currently, SpatialTwin has been deeply integrated with Standard Robots' RoboVerse intelligent collaborative system. In the future, the two parties will integrate their advantages in industrial embodied intelligence and spatial intelligence, empower intelligent manufacturing with digital twin technology, accelerate the large - scale application of industrial embodied intelligence, and jointly build a deep - level ecological cooperation.
Wu Kailiang said that digital twin is a must for smart factories in the era of embodied intelligence. Compared with the high certainty in the era of industrial automation, the autonomy of embodied intelligence determines extremely high computational complexity. Only through infinite pre - rehearsals in the twin world can the output of factories be truly predicted and optimized. GroupCore Technology hopes to work with more ecological partners to promote the in - depth implementation of digital twin technology in industrial scenarios, connect the full - link of intelligent agent decision - making, execution, and optimization, and jointly build a more flexible and efficient intelligent industrial ecosystem.