Step into any automotive welding shop that has been running for over a decade, and you'll likely see robotic arms in four distinct colors from Fanuc, ABB, Yaskawa Electric, and KUKA moving along the assembly line. For a long period, when factories selected industrial robots, the choice was a multiple-choice question with only four answers.
Where the cracks are showing: From a 60% share to a 50-50 split in five years
Around 2020, the "Big Four" families collectively captured nearly 60% of the Chinese market. These four giants have long relied on their accumulated expertise: Fanuc has a closed-loop technical system of CNC, servo, and robotics; Yaskawa's roots are in motion control and servo systems; ABB embeds its robots in a broader automation, vision, and software ecosystem; and KUKA has been deeply involved in automotive manufacturing, heavy-load robots, and automation system integration. Behind the robotic arms they sell are decades of accumulated process packages, customer certifications, engineering experience, and service networks.
However, the data has already carved a clear crack into this old order. In 2025, this change continued to transmit to brand rankings. MIR DATABANK data shows that China's industrial robot market sales exceeded 336,000 units, a year-on-year increase of about 14.2%. Local enterprise Estun Automation shipped approximately 33,400 industrial robots throughout the year, holding a market share of about 10%, and for the first time secured the top spot in overall brand shipments in the Chinese market. Meanwhile, Inovance Technology reached a market share of 9.2% in the same period, ranking third overall and second among local brands, with its SCARA robot share around 25%, continuing to hold the top position. KUKA, the only one of the Big Four to hold its ground, is a wholly-owned subsidiary of Midea, shipping over 32,000 units domestically in 2025, a year-on-year increase of over 30%, with a share of 9.6%. Its China revenue now accounts for around 30% of the entire KUKA Group. Fanuc's fiscal 2025 revenue reached approximately $5.7 billion, up 7.6% year-on-year, hitting an all-time high, with an operating profit margin of 21.4%. Yaskawa Electric's robotics division saw revenue growth without profit growth, with revenue up 4% but segment operating profit down 5.7%. ABB's robotics business recorded revenue of $2.3 billion in 2024, accounting for 7% of the group's total, lower than the group's core segment levels of 16%–18%. Their absolute performance hasn't collapsed; in the first quarter of 2026, Fanuc's China revenue grew 21% year-on-year, with orders surging 55%. It's just that the overall market is expanding faster, and growth cannot mask the continuous decline in relative share.
The most comfortable market structure for the "old money families" was when technological leadership, brand leadership, and scale leadership all held simultaneously. Now, at least the pillar of "scale leadership" is visibly wobbling.
The domestic breakthrough starts with price but has penetrated deep into the supply chain
Reducers, servos, and controllers once formed three hard barriers for domestic robots. Any component constrained by overseas suppliers meant that costs, delivery times, and even the pace of machine iteration were controlled by others. GGII data indicates that the share of domestic RV reducers in the Chinese market was about 41.59% in 2022, rose to 52.89% in 2023, and further increased to 60.80% in 2024. The localization rate for harmonic reducers has climbed from 16.4% in 2017 to approximately 75.11% in 2024. A similar shift is occurring in servo systems. According to data cited in Inovance Technology's annual report from Ruinria Data, in 2024, its general servo system held a market share of about 28.3% in China, ranking first, with Siemens, Panasonic, and Yaskawa at approximately 9.7%, 7.1%, and 7.0%, respectively. This means that domestic companies' grasp of the underlying capabilities of robot motion control can no longer be described as mere "peripheral support."
Whole-machine manufacturers are also moving towards vertical integration. Estun has formed a product chain covering motion controllers, servos, robot controllers, robot bodies, workstations, and intelligent manufacturing systems. Inovance is weaving together servos, control systems, SCARA, six-axis robots, and machine vision. They are replicating part of the method the Big Four used to build their advantage: letting core components, control systems, and complete machines feed each other, rather than simply sourcing parts to piece together a robotic arm.
The localization rate for light-load small six-axis robots has reached 58%, welding robots 62%, and in some scenarios like small and medium-sized auto parts and steel structures, it exceeds 90%. The domestic share for palletizing-specific four-axis robots is over 70%, and collaborative robots have reached 85%. Welding, palletizing, collaborative, and light-load six-axis markets might each look like small battlefields, but combined, they are gradually redrawing the entire market map. The static repeat positioning accuracy of leading domestic light-load small six-axis robots has reached ±0.02 to ±0.03 millimeters, approaching the static indicators of some mid-range Yaskawa and Fanuc products.
A gap is not the same as a breach; what still stands on the city walls
Globally, the Big Four still possess a deep foundation. In 2025, they collectively accounted for approximately 38% of global industrial robot shipments and maintained a clear advantage in high-end applications like automotive manufacturing and semiconductors. The toughest stronghold is automotive vehicle assembly. To this day, high-end production lines at domestic vehicle manufacturers predominantly use robots from Fanuc, KUKA, and ABB. The process inertia and validation systems formed by long-term cooperation are more difficult barriers to overcome than technical specifications. In 2025, Estun secured multiple spot-welding orders on main vehicle welding lines from several carmakers, breaking the domestic record in this field. The door has been opened a crack, but there's still a long way to go before entering the main hall.
The second fact is hidden in customs data: in 2025, the average export price for Chinese industrial robots was about 37,200 yuan per unit, while the average import price was about 97,000 yuan per unit—a gap of nearly double. While the volume has overtaken, the structure remains "selling mid-end, buying high-end." The title of net exporter is genuine, but it marks scale status, not technological status.
The third fact lies in the income statement. Estun, having reached the top, saw record revenue of 4.888 billion yuan in 2025, but its net profit attributable to the parent company was only around 45 million yuan, with a net margin of less than 1%. For every 100 yuan of robots sold, less than 1 yuan is earned. The strategy of trading price for volume has propped up market share but also squeezed profits. The good news is that a turning point has emerged: in the first half of 2026, Estun's net profit attributable to the parent increased more than 21 times year-on-year, with gross margin recovering to 31.8%, and the rising share of high-end products is beginning to pay off.
The most easily overlooked wall is software. Fanuc's ROBOGUIDE, ABB's RobotStudio, and KUKA's KUKA.Sim—these offline programming and simulation software tools have undergone decades of iteration, accumulating a vast library of process packages for mature scenarios. The cost for a customer to switch brands isn't just about replacing equipment; it involves migrating entire production line programs, process knowledge bases, and engineer habits.
Over the past five years, the incremental production lines in China's manufacturing sector have been concentrated in lithium batteries, photovoltaics, 3C electronics, and new energy vehicles. These lines are brand new, without historical baggage. Domestic manufacturers are facing an "incremental examination hall." Estun has worked its way from photovoltaics, lithium batteries, and automotive welding into vehicle factories, while Inovance has moved from servo systems into robot bodies, both following this incremental path. Therefore, domestic robots will likely conquer new production lines, emerging industries, and cost-sensitive markets more quickly, while competition will be prolonged in industries like automotive, which have a large stock of existing equipment and mature process systems.
AI arrives, and industrial robots can't escape economics
An even more interesting variable comes from AI. As humanoid robots become the new focus of capital and industry, the Big Four haven't collectively rushed to "build a human." Instead, they have chosen to equip their mature industrial robots with smarter brains, a choice that reflects the restraint typical of established industrial firms. ABB is combining RobotStudio with Omniverse-related capabilities, Fanuc is promoting the connection of ROBOGUIDE with NVIDIA Isaac Sim, and Fanuc and Yaskawa are participating in NVIDIA's Cosmos physical AI ecosystem, while KUKA is extending towards AI capabilities through its own platform. When factories purchase automation equipment, the decision ultimately comes back to the books: how long the equipment can run, how much cycle time can be improved, how much failure rates can drop, and how long the payback period is. AI entering the industrial site cannot escape this economic logic either.
The shift in the sales rankings is just a ripple on the surface; the deeper change lies beneath: hardware barriers are coming down, and the truly expensive competition is moving towards process, software, global service, and intelligence. The second half of the game isn't about who fights harder, but who first learns to earn money gracefully within others' revenue windows.
Sources referenced: Investment World, Interface News, Beijing Business Today, IT Home