Research on High-End Equipment Manufacturing Sector

Executive Summary

The high-end equipment manufacturing industry in China is currently at a decisive historical turning point, transitioning from an “import substitution period characterized by quantity over price” to an “ecologically mature stage that thrives on intelligence.” As a backbone of the modern industrial system and a core physical carrier of the country’s “new quality productive forces,” its strategic importance has evolved from being a traditional industrial auxiliary tool to a “major national instrument” that plays a crucial role in reshaping global industrial chains. By 2025, research and development intensity is expected to reach 2.8%, surpassing the average level of OECD countries for the first time. This milestone signifies that the industry has moved beyond the benefits of low-end assembly and is now in a phase of high-quality acceleration driven by foundational hard-core technologies and AI algorithms.

Core Business Argument

The greatest market opportunity currently facing this sector arises from the epic resonance between breakthroughs in “artificial intelligence+” foundational technologies and the macroeconomic policy cycle, particularly the large-scale equipment updates supported by long-term special bonds. This resonance not only creates a substantial immediate demand for high-end machine tools, semiconductor equipment, and automated production lines but also accelerates the commercialization and mass production of embodied intelligence technologies, such as humanoid robots, in the long term. The year 2026 is set to be a landmark for the scaling of embodied intelligence, igniting exponential growth from end-to-end neural network control to soft and hard synergy platforms.

However, the sector is also facing critical systemic risks, notably the “valley of death” regarding R&D monetization and the dual pressures of geopolitical supply chains and financial chains. On one hand, extreme technological competition requires companies to maintain high capital expenditures (CAPEX). If they fail to quickly achieve returns on investment (ROI) and create cost-reduction commercial cycles in complex B-end scenarios, they risk running into cash flow crises. On the other hand, vulnerabilities in core foundational components, such as high-precision reducers and industrial control chips, persist, compounded by rising debt costs and financing friction during the green transition. This poses a systemic crisis for hardware assembly manufacturers with weak risk resistance.

Part One: Macro Environment and Industry Chain Overview

1.1 Macro Environment Analysis (PESTLE Analysis)

This study conducts a panoramic scan of the underlying variables affecting the fundamentals of the high-end equipment manufacturing industry based on the MECE principle, using the PESTLE framework to identify the foundational drivers and potential pitfalls influencing industry growth.

Political Factors

The political cycle and top-level policy design are absolute core pillars for the explosion of the high-end equipment sector. The 2026 government work report from the Two Sessions explicitly places “developing new quality productive forces” and “strengthening major technological equipment breakthroughs” at the forefront of its agenda, setting a hard target for annual R&D expenditure growth of over 7% across society. This national strategic commitment provides policy support for long-term basic scientific research. Moreover, innovative financial tools have played a catalytic role. The issuance of long-term special bonds, specifically aimed at supporting large-scale equipment updates and traditional industry upgrades, has decisively reversed the declining investment willingness among downstream manufacturing firms, injecting substantial and assured financial resources into high-end machine tools and engineering machinery sectors.

Economic Factors

As the macroeconomy transitions towards high-quality growth, new quality productive forces have become the strongest engine driving economic expansion. In the first quarter of 2026, macroeconomic data showed a significant 12.5% year-on-year increase in the value added of high-tech manufacturing above designated size, which rose to 16.9% of the total industrial value added, directly contributing 2 percentage points to overall industrial growth. Concurrently, investment in equipment and tools surged by 13.9% year-on-year, while the value added in the integrated circuit manufacturing industry skyrocketed by 49.4%. These hard data points provide compelling evidence that the capital elements of the macroeconomy are undergoing profound structural shifts—moving irreversibly from traditional real estate and infrastructure investment towards high-value added, high-technology high-end equipment and intelligent manufacturing capacities. This counter-cyclical influx of capital provides ample growth opportunities for sector players.

Social Factors

The irreversible transformation of demographic structures and the pressing need for “automation” due to labor shortages create an undeniable long-term logic for equipment automation. With the continuous rise in wages for skilled workers in manufacturing and a plummeting willingness of young people to work in factories, companies are no longer faced with the question of “whether to automate,” but rather “to upgrade or face extinction.” This rigid gap compels end customers to radically change their evaluation criteria for intelligent equipment, shifting from merely considering procurement costs to rigorously assessing the efficiency of replacing human labor, fault-free operation times, and the ROI over the entire lifecycle.

Technological Factors

The leap in technological paradigms serves as the decisive variable triggering the second growth curve in this sector. The ongoing push for “artificial intelligence+” has transitioned from conceptual advocacy to commercial-scale production. Breakthroughs in foundational algorithms, end-to-end neural network control, pure visual spatial perception, and large model multi-modal interaction have transformed high-end equipment from being “blind executors” reliant on pre-configured codes into “embodied intelligent agents” capable of environmental understanding, autonomous decision-making, and flexible execution. Industrial robot production is projected to experience a 28% explosive growth by 2025, directly reflecting the empowerment of capacity through technology.

Legal Factors

As the industry enters a phase of technological depth and global competition, the legal environment has become highly complex. On one hand, core patents—particularly in motion control algorithms and precision reducer design—have become lethal weapons for leading companies to eliminate competitors, leading to an exponential increase in patent litigation surrounding foundational technologies. On the other hand, as intelligent equipment becomes fully networked and equipped with high-precision sensors, the requirements for industrial data security, privacy protection, and compliance with cross-border data transfer are becoming increasingly stringent. Some small and medium-sized innovative enterprises lacking compliance frameworks may easily miss strategic opportunities by triggering compliance red lines in their overseas supply chain engagements or partnerships with foreign clients, making legal risks a hidden barrier to global expansion.

Environmental Factors

Environmental considerations drive a surge in new energy equipment, such as wind turbines, which saw a production increase of 30.1% in the first quarter, and lithium battery equipment, which rose by 40.8%. However, these same considerations pose significant transformation challenges for traditional heavy equipment manufacturing. Under the strong constraints of the “dual carbon” goals and ESG rating frameworks, green financial policies should ideally guide industrial upgrades; however, the actual implementation often faces friction. Some companies, unable to bear the short-term profit erosion caused by environmental upgrades, may choose to obscure environmental information or reduce transparency, which in turn prevents financial institutions from accurately pricing environmental risks and leads to higher loan rates and risk premiums. This debt cost increase, caused by information asymmetry, undermines the effectiveness of green finance support and becomes a real funding obstacle for some mid- to low-end equipment companies seeking advancement to high-end status.

1.2 Industry Chain Mapping and Value Distribution

High-end equipment manufacturing is not merely about assembling physical parts; it represents the efficient circulation of technology, data, and capital across three layers of the industry space. This report breaks it down into three major segments: upstream, midstream, and downstream, with precise positioning of value points.

Upstream Segment: Infrastructure, Advanced Materials, and Core Components

Value Distribution Characteristics: This segment exhibits a funnel shape characterized by high R&D investment, long verification cycles, and extremely high gross margins. It is located at the leftmost peak of the industry smile curve and forms the physical foundation that determines the performance ceiling of downstream equipment. Key areas include precision reducers (RV and harmonic), high-performance servo motors and drive systems, high-performance industrial control chips, high-fidelity visual/tactile sensors, and specialized lightweight alloy materials. In the overall cost structure of machinery, upstream core components often account for 60% to 70% of the total cost.

Choke Points: Despite China’s advantage in overall production, challenges remain in ultra-high-precision reducer design theory, wafer fabrication processes for high-performance industrial control chips, and the core industrial software supporting high-end five-axis linkage. There are ongoing issues with the coordination and transformation between academia, industry, and application, marking a “last mile” bottleneck. This leaves the industry long constrained by technological blockades from overseas giants like Fanuc and Nabtesco. Overcoming these choke points is a life-or-death battle for achieving absolute autonomy in the industry.

Midstream Segment: Core Products, Equipment R&D, and System Control Services

Value Distribution Characteristics: This segment features soft-hard decoupling, system integration, and platform empowerment. Midstream players mainly include providers of industrial automation control systems, high-end CNC machine manufacturers, core semiconductor equipment suppliers, and manufacturers of embodied intelligent robots (both humanoid and quadrupedal). The simple assembly of hardware is rapidly commodifying, with gross margins continually under pressure. The true value has shifted to how to construct a three-layer integrated architecture of “brain (AI models/algorithms), cerebellum (motion control), and body (mechanical hardware).”

Profit Pools: The largest profit pool in the current industrial chain has irreversibly tilted towards “foundational industrial intelligent control platforms” and “integrated soft-hard solutions.” For example, companies that can provide cloud-edge-end deployment capabilities and integrated control platforms, like the new-generation X5 intelligent platform with over 500 foundational interfaces, and have control over the foundational operating system discourse, can not only capture high software licensing and service premiums but also firmly bind downstream customers through platform ecosystems, seizing the most lucrative and sustainable profit shares in the industry chain.

Downstream Segment: Application Scenarios, Process Engineering, and Delivery Channels

Value Distribution Characteristics: This segment is characterized by high customization, capital intensity, and fragmented scenarios. Downstream includes 3C electronic assembly, automotive and parts manufacturing, semiconductor wafer foundries, and warehousing logistics, and is expanding into all-encompassing scenarios such as hazardous materials inspections, commercial services, and even medical rehabilitation. The value in this segment lies in possessing vast amounts of real industrial data and extreme process know-how.

Interactive Logic: The downstream is no longer merely a purchasing party but has become a “testing ground” for deep co-creation with the midstream. Central enterprises, state-owned enterprises, or leading companies with extensive application scenarios have taken the lead in opening up scenario validation rights, driving a crucial leap from “laboratory prototypes” to “production line scaling.” In other words, whoever can first establish a closed-loop on production efficiency and ROI in real downstream workflows will gain ultimate market pricing power.

Part Two: Market Size Estimation and Core Competitive Moats

2.1 Market Size Estimation (TAM/SAM/SOM) and Driving Logic

Based on the latest macroeconomic data from the National Bureau of Statistics for the first quarter of 2026, forward-looking assessments disclosed by the industry, and the calculation models of mainstream investment research institutions, this report conducts rigorous projections of the market size for high-end equipment manufacturing and its core incremental tracks over the next five years (2026-2030).

Core Calculation Logic and Key Assumptions:

1. TAM (Total Addressable Market) Fundamental Driving Logic: Counter-cyclical macro reset. The estimation assumes that the certainty of funds from “long-term special bonds” will continue to be released over the next 3-5 years, supporting a massive demand for upgrading traditional industries. Considering the first quarter’s equipment procurement investment achieved a rapid growth of 13.9%, along with a stringent annual R&D investment requirement of no less than 7%, the overall high-end equipment market is expected to progress steadily at a compound annual growth rate (CAGR) of approximately 11.2%. China’s substantial heavy asset industrial base serves as a ballast for this massive foundation.
2. SAM (Serviceable Available Market) Structural Fragmentation Logic: The tipping point of intelligent penetration. The fast growth of SAM (CAGR ~18.9%) stems from a dramatic internal structural differentiation. The report assumes that as the “artificial intelligence+” initiative achieves commercial-scale application in key industries, traditional low-end devices will be wholly replaced by intelligent equipment. Referencing the first quarter’s phenomenal 49.4% increase in the integrated circuit manufacturing industry’s added value and a 28% surge in industrial robot production, the demand for high-end chip manufacturing equipment and precision intelligent control platforms is on the verge of explosion, forming the core momentum for SAM expansion.
3. SOM (Serviceable Obtainable Market) Exponential Growth Logic: Transition from 0 to 1 in commercialization. The year 2026 is set to be the absolute year for mass production of humanoid robots and embodied intelligence. According to TrendForce data models, global shipments are expected to exceed 50,000 units (an increase of over 700%), with China occupying over 80% of production capacity (conservatively estimated between 28,000 and 62,500 units, or optimistically 100,000 to 200,000 units). This model assumes the domestic main model’s unit price will be in the range of 200,000 to 300,000 RMB, and as the domestic production rate of core components (especially planetary roller screws and dexterous hands) surpasses the breakthrough threshold and scale effects manifest, unit costs will decline by more than 20% annually, ultimately driving the SOM market to achieve an explosive growth rate of over 60% within five years.

2.2 Core Competitive Moat Deconstruction

Based on the reconstructed Porter’s Five Forces model, the traditional logic of “scale is justice” has been completely overturned in the current “AI+ manufacturing” era. The core elements for players in this sector to establish strategic moats are prioritized as follows:

1. First Level (Absolute Barrier): Unique data assets and foundational soft-hard synergy architecture. In the era of embodied intelligence, the physical body of the robot serves merely as a carrier; algorithms and data are its soul. The highest moat lies in whether a company can independently construct a fully integrated architecture of “brain (cognition), cerebellum (motion control), and body.” The first to seamlessly integrate products into complex real workflows (such as injection molding and precision assembly) can acquire exclusive, high-frequency, multi-modal process execution data. This real-world data feeds into end-to-end neural networks, creating a “data flywheel” that makes it virtually impossible for latecomers to surpass in the short term.
2. Second Level (Deep Water Moat): High customer switching costs and binding industry ecosystems. High-end industrial equipment is a production tool that directly affects the profitability of downstream customers. Once a company’s control system or integrated smart equipment successfully integrates with an enterprise’s MES/ERP system and changes workers’ operating habits, the replacement costs become extremely high. Furthermore, by leading the establishment of “innovation alliances” and integrating into the supply chains of Fortune 500 companies (serving over 15,000 clients), leading enterprises have formed “interconnected interests,” effectively locking out potential new entrants.
3. Third Level (Cost and Safety Barrier): Self-research rates of core components and extreme supply chain integration. Given the uncertainties posed by geopolitical factors and tariff barriers, control over the upstream supply chain is crucial for a company’s survival. Companies achieving high self-research rates in choke points such as servo drives, high-precision sensors, and motion controllers, or even controlling five-axis linkage processing centers for high-precision component manufacturing, eliminate supply risks while maximizing their product’s “load-to-weight ratio” (e.g., exceeding the industry average of 0.67) and yield rates. This extreme cost advantage based on vertical integration is the strongest trump card against price wars.
4. Fourth Level (Defensive Moat): Early-stage technology patent clusters and authority in setting industry standards. Although rapid technological iterations shorten the half-life of individual patents, the “patent jungle” surrounding precision transmission structures, integrated control mainboards, and specific industrial applications of AI models remains an effective tool for delaying competitors’ R&D progress. Additionally, participation in or leading the formulation of national/international standards for machine vision and intelligent manufacturing communication protocols significantly enhances brand power, granting exclusive advantages in bids.
5. Fifth Level (Infrastructure Barrier): Compliance qualifications and the ability to capture policy dividends. In the context of new quality productive forces, acquiring provincial-level “specialized and new” qualifications, meeting ESG low-carbon credit thresholds for green finance, and enjoying one-stop sci-tech services and R&D subsidies in high-end equipment manufacturing parks are fundamental survival skills for companies to lower overall financing costs and withstand capital winter. However, such barriers are subject to policy fluctuations and cannot be relied upon as core competitiveness in the long term.

Part Three: In-Depth Analysis of Business Models

The essence of business in the high-end equipment manufacturing sector is undergoing a transformative generational leap: transitioning from a traditional “one-time sale of iron lumps” heavy asset manufacturing industry to a knowledge-intensive service industry that “delivers productivity and algorithmic data continuously through smart hardware.”

3.1 Triple Innovation Reconfiguration of Value Chain

Value Creation: Transitioning from “mechanical definitions” to “software/AI definitions.” Traditionally, value creation relied heavily on materials mechanics and mechanical structure assembly. Today, the origin of innovation is positioned in virtual space. Companies utilize digital twin technology and high-fidelity simulation systems to conduct millions of stress analyses and gait algorithm trials in the cloud. The core value of products is no longer determined by the assembly of steel and motors but by their embedded computational platforms (e.g., 3352 TOPS large model computing power), the number of open foundational interfaces (e.g., over 500 interfaces), and the “systemic adaptive capabilities” provided by algorithms. The measure of value is directly quantified in terms of efficiency in replacing human labor and capacity enhancement rates.

Value Delivery: Breaking through layers of agents, moving towards “scenario co-creation and turnkey delivery.” The traditional sales model relying on multi-level distribution channels aimed at stockpiling is collapsing. Modern high-end equipment, due to its extreme complexity, has evolved the value delivery path into deep direct sales from “expert to expert.” Manufacturers no longer provide cold, bare machines but deeply bind with demand-side needs to offer “turnkey” integrated solutions that include computational power, algorithms, and processes, addressing pain points like “drone blade sorting and precision detection.” Through cloud-edge-end architectures, equipment health monitoring and OTA upgrades have become the new normal in value delivery.

Value Capture: Moving away from one-time buyouts to developing long-tail “money printing machines.” Due to increasing hardware competition leading to lower margins, leading companies have shifted their profit logic to an evolved version of the “razor and blades” model. By deploying smart devices at cost or minimal profit to secure line positions, they then generate recurring high-margin cash flows through software licensing fees for foundational control platforms, subscription-based model inference computing services (MaaS, Manufacturing as a Service), and predictive maintenance services.

3.2 Typical Profit Model Analysis and Market Validation

For enterprises with varying lifecycles and resource endowments within the sector, the following four types of successful commercial models can be summarized:

Model 1: “Vertical Scene Domination” Model Based on Full-Stack Self-Development

Underlying Logic: Companies exercise extreme restraint against blind expansion, focusing intently on a few high-frequency, high-difficulty industrial scenarios. By achieving 100% self-research on core components (reducers, controllers, visuals), they eliminate the cost black hole created by layered supplier markups. They then directly introduce highly compatible embodied intelligent hardware products into client production lines, using aggressive cost data to penetrate client procurement defenses.

Validation Case – Topstar: As a prime example of this model, Topstar has focused on manufacturing for nearly 20 years. They have eschewed flashy features in favor of addressing the harsh environments of injection molding, launching China’s first application-level intelligent humanoid robot, which achieves fully automated sorting and inspection. Their business model has led to astonishing profit elasticity: with revenues reaching 2.51 billion RMB in 2025, net profit attributable to the parent company surged by 130.12% year-on-year to 73.87 million RMB. This “full-stack self-research + scenario closure” strategy retains nearly all profits within the company.

Model 2: “Horizontal Ecosystem Empowerment” Model Driven by Core Platforms

Underlying Logic: Instead of competing with downstream for terminal hardware forms, companies focus on controlling the “throat” that all intelligent equipment relies on—automation control bases and drive neuron systems. By refining inverters, servo systems, and high-end PLCs into standardized platform products, they leverage scale effects to undercut competitors’ costs and, through a “building blocks” approach, provide customized electronic control solutions across dozens of vertical industries such as elevators, new energy vehicles, and lithium batteries, earning excess monopoly profits at the ecosystem platform level.

Validation Case – Inovance: With its integrated electromechanical-hydraulic control technology, Inovance has built a vast industrial ecosystem. While it does not manufacture automobiles, it has become an indispensable invisible leader in the new energy vehicle sector through its electric drive control solutions. 2025 financial reports show total revenues reaching an impressive 45.11 billion RMB (a 21.77% increase), with its dual main businesses of industrial automation and new energy vehicles accounting for a staggering 94.38% of total revenue, alongside significant cash dividends of 1.35 billion RMB, perfectly validating the capacity of its foundational platform model to withstand market fluctuations.

Model 3: “Agile Innovation and Ultimate Cost-Performance” Model

Underlying Logic: This model introduces internet-driven rapid iteration and first-principle thinking. Targeting research or specialized equipment markets long dominated by international giants, it drastically reduces redundant components and leverages China’s comprehensive foundational supply chain to bring products originally priced in the millions down to tens of thousands or even thousands. By achieving extreme cost-performance ratios, they create a substantial long-tail market that previously did not exist.

Validation Case – Unitree: As a leader in the quasi-IPO realm, Unitree has precisely entered the high-dynamics quadrupedal and bipedal robot sector. It has avoided the pitfalls of pursuing overly large system integration by focusing on optimizing key pain points such as robot speed and impact balance. Thanks to its astonishing low-cost mass production capability, its H1 and other products have thrived in domestic special inspections and academic research markets, while establishing absolute category recognition in overseas markets, achieving a dual impact of scaling and high growth.

Part Four: Benchmark Enterprise Case Studies

To clearly deconstruct the survival game rules of different players within the current sector, this section selects two representative benchmark enterprises: one a hundred-billion-level platform leader, and the other an embodied intelligence breakthrough dark horse, for in-depth analysis.

4.1 Global Ecosystem Leader and Automation Control King: Inovance

Core Positioning: The “absolute chain master” of China’s high-end industrial automation foundation. It firmly occupies the core midstream positions of controllers and drive actuators in the industrial chain, with its moat built upon profoundly deep electrical electronic core algorithms, industry know-how data accumulated from over ten million units of historical installations, and extremely high switching costs due to deep binding with leading companies across the manufacturing landscape.

Breakthrough Strategy and Differentiated Approach

Inovance’s rise can be viewed as a textbook “encircling the city from the countryside” success story. Initially facing the monopolistic iron curtain of international giants such as Siemens and Yaskawa in the high-end large PLC and servo system markets, it did not confront them head-on. Instead, it astutely targeted the elevator control niche market overlooked by the giants, utilizing high customization services and rapid response mechanisms to accumulate initial capital. The subsequent differentiated approach manifested as a matrix of “universal bases + industry-specific machines.” It decoupled core technologies to form modular capabilities, quickly replicating its penetration into dozens of vertical industries like air compressors, cranes, and lithium battery manufacturing. A critical differentiator was its forward-looking “cross-industry reuse” strategy—applying industrial-grade high-reliability electric control technology to the burgeoning new energy vehicle sector. This “one fish, many eats” strategy not only significantly reduced R&D costs but also created a dual-main business matrix, achieving revenues of 45.1 billion RMB by 2025 and solidifying its dominance.

Future Strategic Direction

Based on its solid shareholding structure (with stable voting rights under controlling shareholder Zhu Xingming) and ample cash flow, Inovance’s future core strategy is expected to focus on “deep diving into intelligence” and “navigating a great maritime era.” On the technology front, it will undoubtedly integrate industrial AI models directly into its next-generation PLCs and edge controllers, evolving from providing “execution components” to offering intelligent cores capable of autonomous optimization and predictive maintenance. On the market front, facing the slowing marginal growth of domestic manufacturing, it will rapidly localize capacity and technical service networks in emerging markets like Southeast Asia and Eastern Europe, embarking on a strategic expedition from “China’s import substitution king” to “global automation standard bearer.”

4.2 Commercialization Leader in Industrial Embodied Intelligence: Topstar

Core Positioning: The “full-stack self-research implementation expert” in industrial full-scene embodied intelligence. Positioned in the midstream of the industrial chain for equipment manufacturing while deeply penetrating upstream core components, its moat is built on the iron triangle of “millisecond-level integrated control platform (X5) + extreme load-to-weight ratio hardware + real industrial scene mass production data.”

Breakthrough Strategy and Differentiated Approach

Amidst the hype of capital speculation surrounding humanoid robots and the dominance of large models, Topstar has adopted a highly pragmatic and incisive approach of “reverse engineering from scenarios and penetrating deep waters.” Firstly, it rejected the notion of showcasing robots performing flips, instead targeting the manufacturing industry’s most painful issues—harsh environments and high turnover rates associated with injection molding processes. The launch of China’s first humanoid robot “Xiaotuo,” specifically designed for the injection molding industry, has practical service indicators: a single arm weighing only 15kg yet capable of carrying 10kg (with a load-to-weight ratio of 0.67, far exceeding the industry average of 0.2-0.3), combined with a repeat positioning accuracy of ±0.05mm, effectively solving the challenges of narrow spaces and dexterous operations in factory applications.

Secondly, regarding supply chain and cost control, Topstar has implemented an “internally driven hard-core strategy.” The complete self-research of core components (controllers, servos) not only avoids dependency but also directly resolves the high-precision processing and waste loss issues of complex structures like robot legs and joints through its holding subsidiary Evermi, which provides five-axis linkage machining centers, locking in cost advantages at the physical foundation level. Lastly, it exhibits remarkable financial engineering in capital operations. Plans for an H-share issuance by the end of 2025 aim to build a dual financing platform of “A+H.” This move not only broadens its international perspective but also raises abundant low-cost strategic funds for recruiting top AI talent globally and establishing a worldwide sales network.

Future Strategic Direction

As a leader in the head camp that has achieved a commercial closed loop, Topstar’s future strategy will transition from “explosive breakthroughs in injection molding” to “full-dimensional ecological expansion.” Its next-generation X5 intelligent control platform, which has opened over 500 foundational interfaces and features 3352 TOPS computing power, demonstrates its ambition to evolve from merely being a hardware vendor to becoming a “operating system” provider in the era of embodied intelligence. It is anticipated that the next steps will leverage its existing base of 15,000 clients to rapidly replicate the humanoid system validated in injection molding across broad scenarios such as automotive parts assembly, flexible manufacturing in 3C electronics, and inspections in high-risk environments. Additionally, through deep joint ventures with major model giants like Zhipu Huazhang, it aims to enhance its environmental adaptability and autonomous decision-making capabilities, solidifying its dominance as a benchmark for “Chinese embodied intelligence innovation companies” by 2026.