Hangzhou's "Six Little Dragons" and "Eight New Steeds" once wrote a legend on Zhejiang's sci-tech innovation map, and now Tongxiang has finally welcomed its own "AI squad." Drawing on years of industrial strength built up through the World Internet Conference Wuzhen Summit, Tongxiang has officially launched the "Wuzhen Seven" 鈥?Shangyuan Zhixing, Xinshengji Intelligence, Shuhan Technology, Sainad, Dayan Technology, Freetech, and Xinhan Intelligence (listed in no particular order). Many will ask: why these seven? To understand the "Wuzhen Seven," one must abandon the habit of a "checklist-style review." A single thread cannot make a rope, and a single tree cannot make a forest. These seven companies are not a simple assembly of star names. They appear to belong to different industries, operate on independent tracks, and serve different scenarios, yet placed on the same industrial map, artificial intelligence is the main thread running through them all, and a clear AI industrial loop emerges 鈥?from underlying computing power cooling, simulation data, and industrial brains, to embodied robots, intelligent chassis, and full-vehicle smart driving.
In the Xinhan Intelligence laboratory, at the very front of this chain, Xinhan Intelligence plays the role of an AI computing power "cooling specialist." Currently, the power consumption of high-performance computing chips is surging, and traditional air cooling can no longer withstand the heat of high-density computing. Once overheated, expensive computing power will be throttled and lost. This company, founded by a team of master's and doctoral graduates from Zhejiang University, focuses on two-phase phase-change liquid cooling, using liquid vaporization to carry away massive amounts of heat, "cooling down" chips and ensuring stable output of computing power.
With stable and cool computing power, AI still needs massive amounts of high-quality data for training, and what Dayan Technology does is precisely the "virtual training ground" for embodied intelligence. Its self-developed 4D world reconstruction and controllable diffusion model, paired with China's first force-touch interactive glove Shadow Gauntlet, allows robots to "feel" the real world.
At the Sainad Tongxiang production base, the chain continues forward, and intelligence begins to step out of the simulation environment and become physical robots capable of doing hands-on work. Sainad targets the toughest "last 20 meters" loading and unloading challenge in the logistics industry. Its iLoabot series embodied intelligent loading and unloading robots have a single-unit load of up to 80 kilograms and can handle over a thousand boxes per hour. They have already been deployed by leaders such as SF Express and Sinotrans, and the heaviest, most tedious physical labor in warehouses is now being handed over to "digital blue-collar workers."
In the Xinshengji cleaning robot production workshop, also on the embodied intelligence track, Xinshengji has chosen the commercial cleaning sector, which is closer to public spaces. Its commercial cleaning robots rely on physical AI integrating 3D LiDAR and vision, and have already entered more than 40 countries and regions worldwide. Singapore Changi Airport has over a hundred units running year-round, and UK-based Tesco deployed 600 units in one go, setting the largest-scale overseas delivery record for Chinese commercial cleaning robots to date.
On the Shuhan Technology online center interface, when AI enters the production line, the key is to be "useful." On one side are robots working in factories, warehouses, and terminal buildings; on the other side, the "Wuzhen Seven" have driven intelligence onto the road and into workshops, serving the vast real economy. Shuhan Technology installs a thinking "industrial brain" into traditional factories, connecting its self-developed control large model to chemical, environmental protection, and energy production lines, taking over equipment such as fans, valves, and pumps. It does not merely monitor and alarm, but autonomously predicts changes in operating conditions, captures lagging risks in advance, and autonomously adjusts process parameters, leading high-risk, high-energy-consumption heavy industry into an era of intelligent automatic control.
At Shangyuan Zhixing's intelligent unmanned vehicle site, the company has deeply cultivated low- and medium-speed autonomous driving skateboard chassis. Its self-developed i-Skate full-line control chassis transmits braking, steering, and signals entirely through electrical signals and features multiple safety redundancy designs. Its L4 autonomous vehicles have accumulated 100 million kilometers of driving mileage. The modular chassis can quickly transform into urban delivery vehicles, sanitation vehicles, park shuttle vehicles, and security patrol vehicles, and has been deployed in more than 300 cities nationwide.
Freetech, meanwhile, targets two intelligent driving scenarios 鈥?passenger vehicles and commercial vehicles 鈥?and is a leading player in China's independent smart driving solutions. Rooted in its Wuzhen intelligent manufacturing base, it has fully self-developed the ODIN smart driving platform, covering everything from ADAS assisted driving to urban NOA high-level autonomous driving. It has more than 330 mass-production designated projects and deep cooperation with dozens of mainstream automakers, with millions of vehicles cumulatively equipped with this solution on the road.
Tongxiang Phoenix Lake. The "Wuzhen Seven" are not seven isolated star companies, but an AI industrial cluster that a Jiangnan water town has spent more than a decade turning the traffic of the Wuzhen Summit into something rooted in the real economy and capable of sustained growth. It must be said that Tongxiang has adopted a different posture this time. The emergence of the "Wuzhen Seven" is a concentrated presentation of the phased achievements of Tongxiang's artificial intelligence industry development, a vanguard for enhancing Tongxiang's overall innovation capacity, and the starting point for large-scale implementation in the next stage.
Tongxiang is one of Zhejiang's 45 major industrial counties (cities and districts), the province's ninth county-level city with a trillion-yuan output value, and ranks 22nd among China's top 100 counties. It is an "important piece" on Zhejiang's economic map. A city's development cannot rest on past achievements, nor can it rely only on a few top-tier enterprises; it needs more and more "sci-tech innovation stars." When high-quality startup companies continue to emerge across tracks such as chips, computing power, hardware, robotics, aerospace, and biology, forming a pattern of relay growth and complementary coexistence, that is a truly mature sci-tech innovation industrial ecosystem. As the AI industrial wave surges in, Tongxiang, a prepared city, is proving in its own way that county-level economies can not only embrace AI, but also find their own place in the AI era. The second half of artificial intelligence belongs to players that can take root in industry and solve real problems. Tongxiang, a strong industrial city, has already stepped onto the field.
Comments