"Hangzhou has Alibaba, Shenzhen has Tencent, what does Guangzhou have?"

In 2012, a major discussion on the internet economy pushed Guangzhou into this question. The "Millennium Commercial Capital" boasts the highest online shopping volume in the country, yet lacks a leading e-commerce giant. In the history of internet development, Guangzhou clearly lags behind cities like Beijing, Shanghai, Hangzhou, and Shenzhen.

The rise of a certain region is driving Guangzhou's internet economy to overtake on a curve. In 2015, the Pazhou Internet Innovation Cluster in Haizhu District, Guangzhou, broke ground. Though somewhat "late to the party," the "Pazhou speed" is impressive.

By 2022, Pazhou had gathered over 50 leading or headquarters companies such as WeChat, Alibaba, Toutiao, Vipshop, and iFlytek.

The adjacent headquarters buildings make the industrial internet density in the Pazhou pilot zone leading nationwide. In 2022, the Pazhou pilot zone had 1,608 "four-above" enterprises, 345 high-tech enterprises, and 98 headquarters enterprises, far surpassing similar provincial economic development zones in the city, with some indicators reaching or exceeding the industrial density level of Zhongguancun Demonstration Zone.

"'During the 14th Five-Year Plan period, the digital economy enters the 'second half' with industrial internet as the main growth point," said Wu Xiaoying, deputy director of the Haizhu District Science, Technology, Industrial, Commercial and Informatization Bureau, in an interview with Jiemian News.

Currently, the Pazhou industrial internet ecosystem has formed several characteristics: first, platform economy forms comprehensive advantages, with 13 billion-yuan-level platform enterprises in the district; second, digitalization of industrial clusters accelerates, nurturing 9 municipal-level industrial cluster digital platforms such as the Baidu Paddle AI empowerment center; third, rapid growth in vertical fields, with internet information services (advertising), e-commerce, smart healthcare, etc., achieving double-digit growth.

From "Siberia" to New Growth Pole

Heading east along Yuejiang West Road, passing the Hunt De Bridge, geometrically shaped skyscrapers outline Pazhou's new skyline.

Here, separated from the Zhujiang New Town CBD by just the Pearl River, the Canton Fair twice a year boosts the commercial value of this area. In Guangzhou's master plan, Pazhou is positioned as the "Internet Innovation Cluster." In the future, giants like Tencent and Alibaba will engage in face-to-face competition here.

But those unfamiliar with Pazhou's past would be surprised to learn it was once the "Siberia" of Guangzhou's central area.

On the narrow and long Pazhou Island, its ancient port of a thousand years was once the starting point of the Maritime Silk Road. For nearly a century, as an urban-rural fringe area, Pazhou mainly relied on growing sugarcane and fruits for income and was labeled economically backward. After the reform and opening up, factories like Pearl River Brewery, Wuzhou Fan Factory, and Pazhou Refinery brought new revenue, but low land use efficiency and extensive development models always constrained economic growth.

The turning point came after the millennium. In 2001, the Canton Fair venue moved to Pazhou Island, and a year later, the Pazhou Exhibition Center opened. Pazhou was positioned to develop into a new urban center component dominated by exhibitions, international business, and information exchange, also featuring high-quality residential functions.

In 2015, the western part of Pazhou was planned as an internet innovation cluster; in the same year, four giants—Tencent, Alibaba, Fosun, and Gome—invested 78 billion yuan to acquire seven plots in western Pazhou. According to the "Pazhou Internet Innovation Cluster Industrial Development Plan" released in 2016, by 2025, the Pazhou Internet Industry Demonstration Zone will reach a scale of 200 billion yuan, driving the western Pazhou area toward a trillion-yuan cluster.

Currently, Pazhou's positioning has been elevated again to a new industrial highland for the integrated development of "digital + exhibitions + headquarters + high-end commerce," planning to build AI "three calculations" platforms, consumer internet platforms, and industrial internet platforms, leaping to become Guangzhou's new growth pole.

Pazhou is located on the east side of Haizhu District, Guangzhou. Image source: Haizhu Release

Wu Xiaoying told Jiemian Greater Bay Area that Guangzhou and its surrounding deep manufacturing base is driving the development of industrial internet on the production and manufacturing side. The deep industrial foundation and application scenarios attract large manufacturing enterprises to accelerate the upgrading of production and manufacturing sides, gradually establishing independent digital economy entities in Pazhou to empower industry digitization externally, including RootCloud, Guangzhou Pharmaceutical Group, Guangzhou Metro, etc.

In addition, "As the Millennium Commercial Capital and consumption center city, Guangzhou is accelerating the high aggregation of reverse industrial internet on the circulation and consumption sides. The commerce and consumption heritage has given birth to professional industrial internet enterprises focusing on building digital platforms for productive supply and demand services, represented by such as Yigongpin, Baturu; it also fosters the reverse industrial internet path on the consumption side, entering production and supply sides reversely from consumer demand, including Vipshop, Toutiao." Wu Xiaoying said.

"Every industry needs a platform"

In Sany Heavy Industry's No. 18 Factory, intelligent welding robots automatically identify materials for welding in the workshop, heavy unmanned vehicles shuttle full of materials inside the plant, and few human figures are seen among hundreds of robots.

The one "threading the needle" for the equipment is the "RootCloud platform" built by RootCloud with industrial internet operating system as the key. The RootCloud platform can match optimal parameters for each process, machine model, or even each tool based on data feedback from thousands of sensors in the plant, reducing unit manufacturing costs by 29% while expanding capacity by 123% and increasing productivity by 98%.

"Every industry needs a platform." RootCloud co-founder and CEO He Dongdong pointed out to Jiemian Greater Bay Area that nowadays, every industry and enterprise is considering effective paths for digital transformation. Especially in the industrial field, under the background of new industrialization, they are exploring how to organically integrate new technologies such as IoT, industrial internet, and big data with existing automation systems, industrial control systems, and ERP to improve the return on existing investments.

"From the functional role of the industrial internet operating system, first, it achieves interconnection of a large number of industrial equipment, products, and software. Second, it enables rapid development of industrial APPs and widespread reuse of module components."

Guangzhou Pazhou Photo: Zhang Xilong

The core difficulty of industrial internet operation lies in breaking through data between different levels. Currently, the RootCloud industrial internet operating system 4.0 version built by RootCloud can solve data connections at the IoT layer, collect industrial data to the RootCloud industrial internet platform, and make device data, on-site data, and energy data transparent after connection, ensuring production continuity.

It can be said that this unicorn from Pazhou has become a symbol of high-quality development in China's manufacturing industry. So far, RootCloud has provided industrial internet services to over a thousand industrial enterprises in dozens of industrial sub-sectors such as equipment manufacturing, steel metallurgy, automobile vehicles and parts, and electricals; the "Excavator Index" it helped Sany Heavy Industry build covers over 800,000 large construction machines nationwide and is an important indicator of investment activity heat.

Textile Industry's "Ride-Hailing Platform"

Holding a sample fabric, stopping at stall after stall, inquiring one by one, flipping through color cards, and comparing with the naked eye. Ten years ago, Zhijing Technology founder Zhao Zhenhong accompanied an out-of-town friend to Zhongda Fabric Market to find fabric, but after two days, found nothing.

This 85-post computer science graduate thought, could we build a fast and efficient fabric-finding platform where wholesalers and garment factories can trade?

Thus, Zhijing Technology was born.

Compared to RootCloud's PaaS field, Haizhu's industrial internet is more concentrated in SaaS, accounting for over 80%. From the textile and apparel in Zhongda commercial circle to the auto parts city on Xinxin West Road, a large number of traditional industries form the cornerstone of the real economy, while facing urgent transformation needs.

Zhijing Technology is better known to the public as "Baibu," just like Toutiao to ByteDance. Baibu revolutionized the old way of finding fabric: place the Baibu fabric-matching robot on the fabric, take high-magnification photos, and the algorithm analyzes the image and matches the fabric database. In as fast as two minutes, which fabric shops have the corresponding fabric, unit price, size, and material are all displayed on the screen.

The premise is that Zhijing Technology has used fabric-scanning equipment to collect images of nearly 100,000 kinds of fabrics, and AI algorithms extract visual feature information such as texture, color, material, density, and process, combining them into encoded fields to form a unique code for the fabric.

"You can think of it as issuing an ID card to the fabric." Zou Yuanyao, head of Zhijing Technology's brand department, told Jiemian Greater Bay Area that based on ten years of big data accumulation, the recognition accuracy can now reach over 93%.

Zhijing Technology's Baibu fabric-matching robot Image source: Zhijing Technology

Having solved the problem of difficult fabric finding, Zhijing Technology began moving upstream. "In the upstream production environment, domestic factories have insufficient utilization rates, with the industry average around 60%." Zou Yuanyao pointed out.

Just like Didi or Gaode ride-hailing apps linking social vehicles to the platform, Zhijing Technology's "Feisuo Intelligent Textile" aggregates loom capacities. Zhijing Technology Vice President of Full Fabric R&D Liu Yunchun mentioned that factories linked to the platform install AIoT devices on production machines, and the IoT cards inside upload machine production data to the cloud platform. "This way, we can monitor factory utilization in real time and assign large orders to idle factories, just like ride-hailing platforms dispatching to drivers."

From knitting and weaving equipment models and utilization rates across provinces, the "Loom Index" for each time slot, to dyeing vat temperatures and colors, the Feisuo Intelligent Textile platform covers the entire upstream industry chain. Liu Yunchun said that already 9,000 factories and 700,000 looms are linked to the platform. Besides Guangzhou, Zhijing Technology has offices in Shanghai, Hangzhou, Keqiao, Huzhou, and is laying out overseas markets like Southeast Asia.

Compiling the Auto Parts Industry Dictionary

In 2022, Haizhu District, Guangzhou, took the lead nationwide in issuing measures to promote industrial internet development, relying on the Pazhou AI and Digital Economy Pilot Zone to explore new business models such as large-scale intelligent production, networked collaboration, personalized customization, and service extension in fields like home furnishing, automotive, and apparel.

And years before this measure was issued, some enterprises had already standardized the industry by building databases. For example, Baturu, founded in 2013.

Baturu Chairman and Founder Zeng Wan'gui entered the auto parts industry in 1997. "The traditional parts learning cycle is long, with high cognitive barriers but low entry thresholds." He mentioned in an interview with Jiemian News that in the past, doing parts business, "it takes three or four years to figure it out."

In the past, repair shops generally had to go through the following dialogue to determine the parts needed for a car: "What brand is the car? What model? What year? What does the part look like?" Repairing relying on human memory required tedious questioning.

In China, there are over 300,000 car models, each with over 10,000 parts, each part produced by different factories with different names. "So the smallest granularity of parts, i.e., SKU, is on the order of tens of millions."

Starting in 2009, Zeng Wan'gui began compiling the "dictionary" of the auto parts industry—building an auto parts database, defining vehicle models, VINs corresponding to part numbers and Chinese names for all domestic vehicles, including three major libraries: model library, coding library, and name library.

Baturu's "Auto Parts Shop" is the industry's first full-vehicle-parts e-commerce trading platform. Logging into the Auto Parts Shop platform, the page is no different from ordinary e-commerce platforms. After entering the vehicle license number, the page jumps to vehicle model information; then entering the needed part shows part information from different brands and manufacturers, as well as price, warranty period, and delivery time. Next is adding to cart and placing order.

Baturu warehouse shipping dashboard Image source: Baturu

After digitization, auto parts warehouses are marked with clear codes Image source: Baturu

Thereafter, Baturu gradually evolved to output solutions, opening database-based centralized procurement systems, private domain platforms, and distribution systems to customers.

This is also a process of reconstructing the industry chain through digitization. Zeng Wan'gui revealed that the auto parts industry is very fragmented, forming complex intertwined interest chains in circulation over the years.

Through digital platforms, when enterprises conduct business analysis, they only need to pull data to check if merchants or categories have anomalies. Zeng Wan'gui mentioned that in three months, enterprises can achieve sunny, transparent procurement with cost reduction.

In 2009, when Zeng Wan'gui started compiling the first auto parts data, he only intended it for his own chain stores, never expecting it to develop into a platform common to the entire industry.

This is the magic of industrial internet. Compared to consumer internet, industrial internet pushes the boundaries of industrial possibilities outward from production, circulation, consumption, and other links. The "Pazhou Industrial Internet Development Report (2023)" points out that by 2025, China's industrial internet platform transaction scale is expected to surpass consumer internet for the first time, accounting for over 50% of the national e-commerce transaction scale.

In this process, Pazhou is stepping out of Guangzhou, gradually becoming a national industrial internet highland.