The humanoid robot sector is experiencing a significant wave of initial public offerings. YuShu Technology is set to begin subscription on the STAR Market on August 10, potentially becoming the first A-share listed "humanoid robot complete machine" company. ZhiYuan Innovation has initiated its Hong Kong listing process, and other firms like Leju Intelligent and YunShen Tech have had their IPO applications accepted by stock exchanges.

This rush to the capital markets reflects the industry's accelerating development. However, it is important to note that many companies' products remain in early stages, such as exhibitions or small-scale commercial trials. Cases of scalable commercial application with clear profit models are still scarce within the industry. While capital markets can support visionary concepts, valuations ultimately depend on fundamental business performance.

Where to focus efforts

The humanoid robot industry needs to identify precise directions in costs, technology, and application scenarios to accelerate scalable commercial deployment. Currently, the cost of deploying humanoid robots remains high. Beyond the initial procurement of the robot itself, expenses from production line modification, maintenance, and algorithm iteration continue to accumulate, resulting in long payback periods for most scenarios. However, as upstream component localization accelerates and leading companies initiate mass production at the ten-thousand-unit level, the downward trend for hardware costs is already emerging.

Companies can explore innovative business models, such as promoting robot leasing or pay-per-use service models, to lower the entry barrier for downstream customers. They can also rely on scaled delivery to spread fixed costs. To better leverage capital markets, some companies should target specific vertical scenarios, creating replicable and sustainably profitable solutions to avoid homogenized competition.

Technology gap challenge

From a technological perspective, the humanoid robot industry currently faces a structural contradiction where "hardware advances faster than intelligence." However, coordinated efforts in software and hardware are steadily narrowing this gap. China's humanoid robot sector has made continuous breakthroughs in hardware areas like servo joints, motion control, and overall machine structure, enhancing the robot's "cerebellum" capabilities. Yet, the "brain" responsible for autonomous perception, flexible operation, and independent decision-making remains a weakness. Most models can only execute pre-programmed commands, and their ability to autonomously adapt and perform precise operations in complex work environments is insufficient.

To overcome this challenge, more companies should establish a dual-cycle mechanism of "hardware iteration plus scenario data." This involves leveraging real-world factory pilots to continuously accumulate scenario data, which in turn feeds back into the iteration of embodied intelligence large models. It also requires promoting deep collaboration between robot manufacturers and algorithm teams to bridge the gap between simulation training and real-world deployment. Only by advancing both software and hardware can robots transition from "performance products" to reliable "production tools."

Application scenario fragmentation

Furthermore, the fragmentation of application scenarios for humanoid robots has long constrained the industry's commercialization process. Current pilot projects are scattered across fields like automotive manufacturing, warehouse logistics, and cultural tourism displays, but most remain in small-scale trial phases. The operating standards for different production lines vary greatly, making it difficult to quickly replicate a universal solution. Instead of pursuing a robot that covers all scenarios, companies should make strategic choices, focusing on deeply cultivating industries with high standardization, such as automotive, 3C electronics, and logistics. Developing standardized solutions and creating benchmark cases with quantifiable input-output ratios would be more effective. By relying on these benchmark projects, companies can continuously refine their products and then gradually expand their application boundaries.

Long-term outlook

In the long run, the commercialization path for humanoid robots will face multiple challenges. However, by staying grounded, continuously advancing technological innovation, and refining business models, the humanoid robot industry will eventually overcome its bottlenecks and break through to new levels of development.