←News detail

The Beijing NewsUpdated Oct 30 12:44

Xiaoyu Zhizao's Qiao Zhongliang on Embodied Intelligence: Big Tech Competition, Divergent Approaches, and Companies Going Global

AI summary

Qiao Zhongliang, founder and CEO of Xiaoyu Zhizao, said that startups facing competition from big tech should adopt differentiated positioning and go deep into niche, high-barrier tracks such as industrial welding, avoiding head-on competition in red oceans like general-purpose computing power and general-purpose humanoid robots. The company's product, launched in July, has seen core customer repurchase rates exceed 50% in just over two months, and it has set an explicit goal of reaching annual sales of 100,000 units within 3–5 years; it began laying out overseas business in May this year. He predicts that in the next decade, fewer than 10 core companies will remain in the industry.

Body · AI translation

On October 27, the "Industrial Embodied Intelligent Welding and Inspection Technology Joint Laboratory," jointly established by the National Steel Structure Quality Inspection and Testing Center of MCC Building Research Institute Co., Ltd. and Beijing Xiaoyu Zhizao Technology Co., Ltd., was officially unveiled and established. Qiao Zhongliang, founder and CEO of Xiaoyu Zhizao, said that welding is not an ordinary process—every weld seam is tied to engineering safety and human life. Robots can weld seams faster, but if quality does not meet standards, the higher the efficiency, the greater the risk.

Qiao Zhongliang previously worked at Xiaomi for 12 years. In an interview with media including The Beijing News Shell Finance, he said that in the face of competition from big tech companies, the core of a startup's survival is differentiated positioning, deep cultivation of edge high-barrier tracks, and avoiding head-on competition with big companies in red-ocean tracks such as general-purpose computing power and general-purpose humanoid robots. Xiaoyu Zhizao officially launched its overseas business layout in May this year. He expects that in the next decade, fewer than 10 core enterprises will remain in the industry.

Qiao Zhongliang, founder and CEO of Xiaoyu Zhizao. Photo provided by the interviewee

Annual sales of 100,000 units is the hard threshold for building cost and scale barriers

Currently, embodied intelligence, alongside quantum technology and biomanufacturing, has been selected as one of the country's six major future industries in the "15th Five-Year Plan." What deep transformations will embodied intelligence bring to traditional industrial production?

Qiao Zhongliang said that the core reason the country includes embodied intelligence in future industries is its recognition of its core value in reshaping productivity, rather than mere conceptual innovation. The essential difference between embodied intelligence and traditional automation is not simply efficiency improvement, but breaking the century-old bottleneck in which automation only adapts to standardized scenarios, enabling non-standard industrial scenarios such as ships, heavy steel structures, and construction machinery to achieve automated and intelligent operations.

"From deployment data, the equipment we have deployed in more than 20 partner factories currently maintains an overall arc-on rate of 70%, with a peak of 75%, and the product welding qualification rate is stable at over 90%. Operating quality, stability, and efficiency are all superior to manual work," Qiao Zhongliang noted. The industry's overall supply chain and technical system are not yet mature, but the pace of deployment and iteration is continuing to accelerate.

Welding is a global rigid-demand track. What is the latest progress in international business? Qiao Zhongliang introduced that Xiaoyu Zhizao officially launched its overseas business layout in May this year and is currently in the early deployment stage. Welding is a global rigid-demand industry; infrastructure construction, shipbuilding, and heavy machinery are global general tracks, and overseas market space is broad, with global annual demand reaching 100,000 units. The deployment of ToB intelligent equipment has a fixed cycle; from pilot validation, customer certification, and ROI verification to large-scale deployment, it takes four to five years.

In Qiao Zhongliang's view, laying out overseas later would mean missing the track's window period. "Therefore we adhere to a strategy of early layout and high-startpoint overseas expansion, prioritizing developed countries and regions with the strictest industrial standards and the highest technical requirements, honing product capabilities through high-difficulty scenarios, and then achieving global downgraded replication."

The company's product was released in July, and it has been more than two months since then. What is the repurchase rate of core customers? Is there a clear plan for staged shipment targets? Qiao Zhongliang introduced that more than two months after the product's official deployment, the repurchase rate of core customers has exceeded 50%; the company has clearly set the core goal of sprinting toward annual sales of 100,000 units in 3–5 years. This is the hard threshold for the company to enter the industry's endgame and build cost and scale barriers, and it is also the company's core strategic goal for long-term stable development.

In the next decade, fewer than 10 core enterprises will remain in the industry

In the face of big tech companies entering the field and traditional robot companies expanding across sectors into the welding track, how will the company build competitive barriers? In Qiao Zhongliang's view, the fact that many leading enterprises are crowding into the field fully proves that embodied intelligence is the core golden track of the next decade and an infrastructure-level technology for industrial development. He judges that in the next decade the industry will be highly concentrated, with ultimately fewer than 10 core enterprises remaining, and robot intelligence technology will become a general infrastructure for the entire industry, just like the internet and cloud computing.

Qiao Zhongliang has 12 years of management experience at big tech companies. He mentioned that Xiaoyu Zhizao's core differentiation is its pragmatic deployment route focused on non-standard industrial scenarios. In the face of competition from big companies, the core of a startup's survival is differentiated positioning and deep cultivation of edge high-barrier tracks, avoiding head-on competition with big companies in red-ocean tracks such as general-purpose computing power and general-purpose humanoid robots.

"Big tech business unit teams generally have problems such as insufficient strategic resolve, scattered resource investment, and weak determination to deploy. We, by contrast, focus on the non-standard high-barrier industrial welding track, concentrate all resources on saturated attacks, and quickly build scenario advantages, data advantages, and deployment advantages to form our own dedicated base," Qiao Zhongliang believes.

Currently, the valuation heat of humanoid robots is far higher than that of industrial embodied intelligence. Unitree Robotics recently completed its listing, and the valuation gap between the industry's two major tracks continues to spark discussion. Regarding the two major tracks of humanoid and industrial embodied intelligence, Qiao Zhongliang believes that the two are different paths to the summit of general intelligence, like the south slope and north slope of a mountain: the ultimate goal is the same, and each path has its advantages and disadvantages.

English text is being prepared.

The embodied intelligence industry is extremely short of real physical-scenario data, and data has become a hot topic in the industry. Qiao Zhongliang believes that computing power, algorithms, and data are the three core elements of the AI industry. At present, the threshold for obtaining general computing power and basic algorithms is gradually lowering. Real physical-scenario interaction data, especially data on failures, anomalies, and insufficiently covered scenarios, as well as the ability to continuously iterate based on such data, are becoming an important differentiated barrier for embodied intelligence companies. Simulation data and public data can provide basic support, but it is difficult for them to fully cover the complex working conditions of real industrial scenarios. The full-link data accumulated from real factory deployment, especially cases of operation failures, scenario anomalies, and adaptation deviations, is particularly critical for continuously driving model iteration and improving system capabilities.

Preview Live broadcast | Citizens Dialogue with Top Leaders: Deepening the Construction of an International Consumption Center City

2 In high-threshold livelihood scenarios, in what posture should AI enter? 54,000

3 Autumn Scenery in the Capital | Autumn in Ditan Is Growing Stronger, the Ginkgo Avenue Enters Viewing Mode 51,000

4 Xiaoluohao | The split between Audi in the north and south is not the endgame; three questions need continuous answers 36,000

5 Behind the Xiangtan University impersonation case: the family ties, interests, and lawsuits of a mother and her two sons 26,000

Latest

Integrated digital-intelligent platform for major chronic disease prevention, control, and treatment completed Just now

Kitchen appliance sector loses momentum in Q3, net profits of Robam, Xinbao, and Bear under pressure 10 minutes ago

Big-tech competition, route differences, and companies going global: Xiaoyu Zhizao's Qiao Zhongliang again discusses embodied intelligence 11 minutes ago

Driven by markets including China, Coca-Cola's global unit case sales grew 5% in Q3 21 minutes ago

First three quarters' performance flat year on year, Da An Gene enters the "Guangzhou Pharma era" 21 minutes ago

Hot discussion

Hiring assassins to kill "Huijian": eradicating telecom fraud completely | Beijing News quick comment

Dialogue with Sun Jian of BTG Homeinns: Continuing the historical accumulation of a national brand, "Jianguoli" creates a wonderful lifestyle gathering place in Beijing's sub-center

Source

The Beijing News