Ant Group Leads RMB 100 Million Round in Juciao's Fabric-Based Robot E-Skin
Ant Group backs Juciao Industrial in a RMB 100 million Series A, signaling a strategic pivot in China's embodied AI investment cycle toward upstream sensory components—even as overall sector funding contracts sharply.
China's embodied AI funding machine is cooling at the macro level, but a narrow band of upstream component makers is bucking the trend—and tactile sensing is emerging as the most contested terrain.
Juciao Industrial, a Shandong-based fabric tactile sensor startup founded less than three years ago, announced on October 29 the close of a RMB 100 million-plus (approximately US$13.9 million) Series A round, led by Ant Group, with participation from returning investors Zero-Yi Ventures, Weihai Port Rong, and Qingdao Hexin. The deal stands out not merely for its size, but for what it reveals about where sophisticated capital is repositioning within China's humanoid robot supply chain.
The broader sector data frames the context starkly: according to IT Juzi, embodied AI financing in October 2026 totaled RMB 10.105 billion (US$1.40 billion) across 33 deals—down from RMB 11.028 billion (US$1.53 billion) across 60 deals in September. The deceleration is unambiguous. Yet within that contraction, tactile sensors, dexterous hands, and electronic skin components continue to attract disproportionate attention, functioning as the sector's last standing growth pocket.
Funding Logic Shifts From "Smart Bodies" to "Bodies That Feel"
The investment thesis underpinning Juciao's round reflects a structural reappraisal of where the true data bottleneck lies in embodied AI development.
Vision data can be sourced from internet video corpora. Language models feed on text. Tactile data—shear force vectors, material texture gradients, deformation dynamics—cannot be synthesized from existing datasets at meaningful fidelity. Simulation engines struggle to replicate the physics. Every usable data point must be harvested frame by frame through physical interaction. Industry participants note that tactile data commands a price premium roughly two times above equivalent visual data: video data runs in the low hundreds of RMB per hour; tactile data reaches into the high hundreds.
That scarcity premium is already showing up in deal flow. In the first seven months of 2026, China's humanoid robot sensor segment recorded nearly 20 financing events totaling over RMB 3.1 billion (US$430 million), with July alone absorbing approximately RMB 1.56 billion (US$217 million). The implication is consequential: whoever controls a scalable, cost-efficient tactile data acquisition pipeline effectively controls a critical chokepoint in the current embodied AI development cycle.
Three Technical Routes Compete—None Yet Dominant
China's tactile sensing landscape is characterized by high fragmentation across three competing architectures, each targeting distinct anatomical zones of the humanoid form.
Hall-effect magnetic sensing, exemplified by Pasini, targets high-precision multi-axis force detection at fingertip nodes in dexterous hands. Vision-tactile fusion, the approach adopted by Daemon Robotics —which Ant Group also led a round in two months prior—achieves spatial resolutions in the tens of thousands of sensing points per square centimeter, generating image-format output compatible with end-to-end neural network architectures.
Juciao Industrial occupies the third lane: woven electronic skin. Its process architecture departs from conventional layered thin-film pressure-resistive construction, which stacks flexible substrate, pressure-sensitive layer, and conductive layer sequentially. Instead, Juciao integrates piezoresistive material and electrode structures directly into individual functional fibers—using nylon as a base filament—then weaves those fibers into a sensing fabric. External force deforms the conductive pathways within the fiber network; downstream algorithms convert the resulting signals into tactile data.
The geometry of competition is not winner-takes-all. Fingertip and dexterous hand applications demand high spatial resolution and rapid multi-axis response—advantages that currently favor vision-tactile and Hall-effect approaches. Torso panels, mechanical arms, and foot pads require large-area, conformable, breathable, and cost-controlled coverage—a profile where woven fabric holds natural advantages. The industry trajectory most likely runs toward a hybrid, anatomically partitioned architecture rather than convergence on a single solution.
Ant Group's two investments—Daemon Robotics for high-precision vision-tactile, Juciao for large-area fabric coverage—precisely map this bifurcation. The portfolio logic is not a bet on one technology winning; it is a hedge across the anatomical segmentation of a humanoid body.
Juciao's Production Record Provides the Differentiated Thesis
To assess whether the Juciao round represents a genuine sector inflection or opportunistic capital positioning, the company's operational track record merits scrutiny.
Juciao was incorporated in late 2023. Founder Zhang Jingqi holds a degree in electronic engineering from the University of York and previously served as CTO of a Xiaomi ecosystem company; the core team draws from the Chinese Academy of Sciences and major smart hardware manufacturers. The company's roll-to-roll production line, disclosed in May 2025, carries annual capacity of 200,000 units at a reported yield rate of 95%—a figure that, if sustained at scale, addresses one of the core criticisms leveled at flexible sensor manufacturers: manufacturing consistency.
By July 2025, local government reports indicated that Juciao's order backlog extended to October, with first-year production value approaching RMB 10 million. In September 2025, the company marked the shipment of its 1,000th humanoid robot fabric electronic skin unit, claiming roughly 100 domestic and international clients—including Unitree Robotics, Songyan Power, and Zero Power —within fewer than six months of commercial launch.
The revenue architecture is tripartite: humanoid robot electronic skin serves as the valuation anchor and technology credentialing asset; smart eldercare products (deployed at Taikang Community, Guotou Health, and CR Care function as the current cash flow engine, with 2024 orders already exceeding RMB 10 million; automotive smart cabin applications are positioned as the longer-term growth lever.
Structural Challenges Remain Before E-Skin Becomes Standard Equipment
The market trajectory appears compelling on paper. Industry projections put China's humanoid robot electronic skin demand at 1.525 million square meters by 2030, corresponding to a total addressable market of approximately RMB 27.4 billion (US$3.8 billion). Traditional textile manufacturers, precision sensor companies, and advanced materials firms are accelerating entry, intensifying supply-side competition.
Yet the industry remains structurally early-stage. Standardization is the most acute bottleneck: the mechanical architecture of humanoid robots varies substantially across manufacturers and models, forcing bespoke skin tooling and pattern development for each platform. Molds and templates cannot be batch-reused, compressing margins and extending lead times. High-end sensor modules incorporating proprietary packaging, sensing chipsets, and specialty materials carry price points that cost-sensitive OEM integrators are reluctant to absorb. Field durability—signal drift, sensitivity degradation, and lifespan compression under sustained high-frequency bending, industrial particulate exposure, and thermal cycling—remains an open engineering challenge across the fabric route specifically.
Juciao's own product roadmap acknowledges the precision ceiling inherent to woven architecture. Its Tactile Skin 2.0 iteration introduces adjusted weave geometries combined with model-based training to extend perception of force directionality, sliding tendency, and shear loading—but whether the resulting precision profile satisfies the high-end requirements of dexterous hand fingertip applications remains unvalidated at production scale.
The customer concentration risk is equally material: operational dependence on a handful of anchor clients including Unitree introduces revenue continuity uncertainty as the humanoid robot OEM competitive landscape itself remains unsettled.
Capital Is Buying Optionality, Not Confirmation
The Juciao Series A should be read as a strategic option purchase, not a verdict on technological convergence. In a window where tactile sensing architectures have not yet consolidated and the data acquisition bottleneck has not yet been resolved, Ant Group is paying to maintain a position in a player with demonstrated manufacturing capability across a differentiated physical substrate.
The real test for Juciao—and for the broader fabric tactile route—is conversion: whether first-mover process advantages translate into stable delivery performance, modular standardization, and model-compatible tactile datasets at scale. If that conversion succeeds, woven electronic skin could establish itself as the commodity-grade full-body sensing infrastructure for humanoid robots. If it does not, the funding window may compress rapidly as competing approaches mature and better-capitalized entrants absorb the large-area coverage use case.
The moment when robot skin stops being a product category and becomes an unremarkable component—present in every unit, discussed by no one—will mark the sector's true maturity. Capital today is pricing the probability of reaching that moment. Which route gets there first remains an open question.
Related Coverage:
Humanoid Robots Pivot From Novelty to Utility as Big Tech Enters the Fray
RobotEra Founder: Learning from Humans is 'Shortest Path' to General-Purpose Robots