Revenue has declined, but profits have returned! Behind Tusda's turnaround from loss to profit: using "subtraction" to boost earnings. Can embodied intelligence become a new engine?

Recently, Guangdong TOSIDA Technology Co., Ltd. (hereinafter “TOSIDA,” 300607.SZ), an industrial robotics company, released its 2025 annual report. The data shows that the company’s full-year operating revenue was RMB 2.51B, down 12.59% year over year; however, net profit attributable to shareholders of listed companies was RMB 73.8725 million, up 130.12% year over year, successfully turning from loss to profit. This “reduced revenue, increased profit” performance profile reflects that this industrial automation enterprise with nearly 20 years of history is undergoing a profound structural adjustment.

Under a strategy of proactively shrinking low-margin project-based businesses and concentrating resources on industrial robots and embodied intelligence, TOSIDA is trying to answer a question: after “doing less” leads to short-term profitability, where does the next stage of growth come from?

What did the business contraction bring?

Looking through the financial report, the most conspicuous change at TOSIDA in 2025 is the restructuring of its business composition. The company actively reduced its intelligent energy and environmental management systems business; revenue from this segment fell 25.55% year over year to RMB 915 million. This segment used to be an important component of TOSIDA’s revenue, but its gross margin was negative and it consumed substantial capital. In its performance report, the company stated that the business had “basically completed the divestment,” and that subsequent revenue would decline further.

The cost of the contraction is a drop in revenue scale—from RMB 2.87B in 2024 to RMB 2.51B. Meanwhile, the annual revenue share of product-based businesses increased by 6.67 percentage points, and the share of gross profit contribution reached 60%, becoming the company’s main pillar for profits. Gross margins for core product lines—including industrial robots and injection molding equipment—also improved; the overall gross margin rose from 14.59% to 28.25%.

Zhang Xiaorong, President of the Deep Science Research Institute, told reporters from the Huaxia Times that this strategy of “giving up low-margin business for profitability” has a clear effect of financial improvement in the short term. “Although revenue is lower, by cutting unprofitable businesses directly turning losses into profits, the company’s finances are healthier. In the long run, putting both money and effort into the main business makes it more competitive. This approach of sacrificing scale first while preserving profit is quite pragmatic; short-term pains for long-term development is a wise choice for manufacturing companies.”

Zhou Di, a senior engineer at Rongrong Technology and an expert in the National Science and Technology Library of the Ministry of Science and Technology, added to reporters from the Huaxia Times that, in the long run, concentrating resources on core businesses such as industrial robots helps build technological and market barriers. “Although the company faces pressure on scale in the short term, its development will be more focused and more sustainable.”

However, one unavoidable question is: after the intelligent energy business is basically divested, can other businesses support future growth?

TOSIDA told reporters from the Huaxia Times that future growth will be carried forward around the “new decade, three-step plan” strategy: first, “thicken the base,” consolidating right-angle coordinate robots and injection-molding supporting equipment; second, “strengthen major equipment,” deepening the layout of CNC machine tools, injection molding machines, and so on; and third, “make embodied intelligence real,” building an intelligent ecosystem across the full domain.

The performance report shows that in 2025, the company’s robot component unit sales reached 10,437 units, up about 13.7% year over year, but the overall revenue of the industrial robot and automation application systems segment fell 9.24% year over year to RMB 685 million. This means that the increase in unit sales has not yet fully offset the revenue pressure caused by price or structural changes. How to convert unit sales advantages into sustained growth in revenue and profit remains a challenge the company faces.

For the reasons behind the revenue decrease, TOSIDA explained that in the early stage the company’s automation application systems business focused on top-tier 3C customers, and orders and revenue scale from customers in other industries declined; the company has become more focused on R&D and layout for “robot +” applications, improving its standardized production capacity and lowering the share of business involving personalized projects. However, as cooperation depth and breadth with top-tier 3C customers increase, the order scale for related businesses continued to grow. At the end of 2025, the company’s in-hand orders were up 116.64% year over year. In industrial robots, the company’s product competitiveness has continued to improve, and its strategy of targeting major customers has yielded results; process and application advantages have become more apparent. Operating revenue increased year over year, including 25.32% growth for its self-produced multi-joint robots, and 7.35% growth for right-angle coordinate robots; total annual shipments of robot products were about 12,000 units.

The “industrial route” of embodied intelligence

At a time when embodied intelligence has become the hottest track in the market, TOSIDA did not choose the grand narrative of “general humanoid robots,” but instead started from the injection-molding scenarios it knows best. In 2025, the company rolled out a humanoid robot “XiaoTuo” for injection-molding workshops, as well as products including a quadruped robot “Xingzi” and AI flexible sorting workstations.

This strategy is regarded by the outside world as “driving nails to make a hammer.” TOSIDA also pointed out in interviews that it will analyze and deconstruct injection-molding process steps to form general process kits adapted to various industrial scenarios, using the generality of injection-molding scenarios to expand horizontally into more industries. At present, XiaoTuo’s main application scenarios include material picking, palletizing, and packaging in warehousing and logistics, as well as autonomous loading/unloading and sorting during production.

How is this “scenario-first” path different? In Zhang Xiaorong’s view, embodied intelligence is still, overall, “an edifice in the air,” lacking truly grounded scenarios. “TOSIDA starts from the injection-molding workshop, which is more down to earth. It rolls out faster; it can use existing customers for testing and soon get into factories to be used. Commercialization is more stable—factories have real needs and are willing to pay, so it can quickly make money.”

Zhou Di also noted that rapid validation through existing customers keeps costs more controllable and payback faster, meaning “profit certainty is far higher than the general-purpose route.” However, obstacles to cross-scenario reuse of technology—from injection molding to broader industrial scenarios, and even to commerce, services, and homes—cannot be ignored.

TOSIDA acknowledged that the main challenges lie in high data costs, weak adaptability, and insufficient generalization ability. To address this, the company jointly established Matrix Zhituo with Zhipu Huazhang to develop a low-cost, highly adaptable portable gripper data acquisition solution, laying the data foundation for embodied intelligence model training. This is also a key component of the company’s business closed loop of “scenario + product + data + AI.”

In competition for the quadruped robot “Xingzi,” facing companies that have already built layouts such as Unitree Technology and Yunshenchu, TOSIDA’s strategy is to treat it as an extension of its robot product matrix, creating synergistic effects with its “TuoXingJi” embodied intelligence product series, industrial robots, and AI workstations—providing differentiated overall solutions rather than simply competing on individual products.

After profitability, the long run has just begun

In the Guangdong province where TOSIDA is located, the robotics industry is thriving. Data shows that in 2025, Guangdong’s industrial robot output was 336.3k units, accounting for 43.5% of the national total, ranking first nationwide for six consecutive years; service robot output was 15.1821 million units, accounting for 81.7% nationwide. As the largest robotics-industry province in the country, Guangdong has formed a full industrial chain layout covering software, hardware, and core components.

Zheng Lei, Chief Economist of Samoyed Cloud Technology Group, analyzed for reporters from the Huaxia Times that this industrial cluster provides TOSIDA with threefold support: first, extremely fast supply-chain responsiveness—within the Greater Bay Area there is a complete robot supply chain, with “iteration speed 10 times that of Silicon Valley and costs only 1/10”; second, leading scenario richness—Guangdong has all 31 categories of manufacturing industries, and TOSIDA’s 15k existing customers are concentrated in the Pearl River Delta, making testing and data collection costs very low; third, efficient integration of industry, academia, and research—“Robot Valley” in Shenzhen clusters institutions such as Southern University of Science and Technology and the Chinese Academy of Sciences, forming a closed loop of “basic research -成果转化.”

But large-scale, industrialized deployment of embodied intelligence still faces key bottlenecks. The biggest obstacle at present is the “data dilemma.” Zheng Lei believes that insufficient supply of high-quality data includes high acquisition costs, difficulty extracting tacit knowledge, lack of anomalous data, and factories’ “data governance” power barriers. Zhang Xiaorong said that hardware and algorithms remain bottlenecks: “Hardware is expensive, hands aren’t dexterous enough, and algorithms aren’t smart enough; if you switch to another scenario, it won’t work well, and the cost to modify complex factory environments is high.” Zhou Di added that insufficient technical maturity, relatively high costs of complete machines, and an incomplete industrial-scenario adaptation ecosystem are also limiting factors.

For the application outlook over the next 3–5 years, Zheng Lei believes that industrial embodied intelligence will be deployed at scale first in structured industrial scenarios such as material handling, sorting, loading/unloading, and logistics warehousing—environments are controllable, tasks are standardized, and ROI is clear. 2025 is viewed as the “year of landing,” and “semi-autonomous + partial group collaboration” will become an important breakthrough. Zhang Xiaorong offered more specific scenario expectations: injection molding, 3C electronics loading/unloading, automotive assembly, logistics handling, and so on—“these places have simple processes and high repetition, strong need to replace human labor, and are easier to roll out.”

TOSIDA’s 2025 annual report outlines a typical transformation case of a traditional industrial automation enterprise at the intersection of industry cycles and capital heat. By proactively shrinking low-efficiency businesses, it achieves a financial turnaround into profitability; by betting on embodied intelligence, it seeks to secure a foothold in the next wave of technological advances.

But challenges are also clear. The endogenous growth of product-based businesses has not yet fully kicked in, and embodied intelligence’s migration from injection-molding scenarios to a broader market still needs to overcome data and technology barriers. And competitors—whether startups or peers—are accelerating their deployments. TOSIDA has chosen a “not sexy” but solid path: coming from industry, and going back into industry.

Whether this path can work depends on the deployment speed of its “scenario + product + data + AI” closed loop, and its ability to convert the existing customer base accumulated over the past 20 years into the first paying users of embodied intelligence. This long run has just begun; whether profitability achieved through “doing less” can ultimately be transformed into “adding more” style growth will be answered by time.

Massive information, precise analysis—right on the Sina Finance APP