The Fall of Daluo Technology, One Year On: How an Embodied Unicorn Died

Spring 2026 finds the embodied intelligence sector mid-capitalization wave. Unitree Technology pursues an A-share listing; AGIBOT readies for a Hong Kong IPO; Xinghaitu and Fourier Intelligence have concluded shareholding restructurings. Valuations across numerous embodied firms now exceed RMB 10 billion. Capital alignment has become the defining security strategy for embodied companies.

Historically, high valuations have not guaranteed stability. The cautionary exemplar: Daluo Robot, a once-celebrated embodied unicorn.

The embodied empire that founder Huang Xiaoqing spent a decade constructing was once valued above RMB 20 billion. It collapsed abruptly just over one year ago. Who, precisely, "killed" this embodied unicorn?

Consider a counterintuitive fact. Social media widely attributes Daluo Robot's demise to a fractured capital chain, employee litigation, and creditor disputes. In reality, by March 2025, Daluo's total court-enforced obligations stood at just RMB 35.3 million. For a firm founded in 2015 that secured seven financing rounds and once commanded a valuation of RMB 22.3 billion — China's "robot industry unicorn leader" — this sum was immaterial. Its Guangzhou office tower and hundreds of acres of industrial park land were wholly owned assets; liquidation alone could have resolved its distress. Why, then, did the company dissolve, its founder depart to incorporate a new entity in Hong Kong?

The trigger for Daluo's collapse, therefore, originated not downstream but upstream — in founder Huang Xiaoqing's equity structure.

01

The Limits of the First-Generation Embodied Entrepreneur

Chinese embodied-intelligence entrepreneurs can be divided into three generations.

The third generation — post-1985 engineers and scientists such as Unitree's Wang Xingxing, AGIBOT's Zhi Huijun, and Star Dynamics' Chen Jianyu — views algorithms and models as the true moat. For them, robots are not hardcoded control logic but learned motion strategies. Hardware is secondary.

The second generation, by contrast, favors "traditional control." The robot body is treated as a "product," and the primary concern is the "equipment sales" business model.

The first generation of robot entrepreneurs consists largely of those born in the 1960s, bearing the weight of their era's limitations. They tend to view robots as mere actuators, underestimating the complexity of physical interaction, while relying excessively on "relationship-driven" models — government procurement and state-backed financing. Huang Xiaoqing typifies this cohort. Prior to founding Daluo, he was a decorated technology executive with an impeccable resume.

Upon graduation, Huang Xiaoqing specialized in electronic communications, spending five years at Bell Labs — an institution regarded as the temple of communications science. There, he led the development of a novel system that compressed seven unwieldy subsystems into three, earning widespread recognition within the lab.

In 1995, he returned to China briefly, bringing Bell Labs colleagues to UTStarcom before shuttling between Silicon Valley and Hangzhou. China's telecom sector was then slowly entering the 3G era. Sensing opportunity, Huang Xiaoqing founded Wacos, a venture to empower the 3G transition — but the concept proved premature, and Wacos 3G collapsed. In 2007, China Mobile extended an offer: president of its Research Institute. Over eight years at China Mobile, he witnessed the 3G-to-4G transition firsthand.

By 2015, as the 5G era began to sprout, Huang Xiaoqing again sensed opportunity — and change. An entrepreneurial urge emerged, though its shape remained unclear. Before founding Daluo, his entire career had been spent in telecommunications; he had zero background in robotics.

Per Huang Xiaoqing's own telling, the robotics pivot originated from a sci-fi dream. A devoted Star Trek fan, he was inspired by the android "Data" — fusing cloud communications with robotics to formulate the "cloud robot" concept.

Many former Daluo employees, however, dispute that narrative, characterizing Huang less as an innovator and more as a conduit. His frequent cross-ocean work afforded early exposure to Carnegie Mellon professor James Kuffner's "cloud robot" concept, which he imported and repackaged for the domestic market.

02

Robotics' "Communications DNA"

Capitalizing on his telecommunications pedigree and the contemporaneous "cloud intelligence" buzz, Daluo Technology raised funding with remarkable speed.

Daluo closed its Series A in 2016, raising $30 million from SoftBank Group, Foxconn, and others — Hon Hai's Terry Gou then helming Foxconn's parent. SoftBank doubled down the following year, co-investing $100 million with Zhongguancun Development Group, a record in the cloud intelligence space.

Daluo's trajectory was remarkably smooth — it achieved unicorn status within a single year.

In practice, however, such abundant capital did not translate into meaningful product or technology breakthroughs. Daluo's roadmap encompassed multiple robot lines: the XR series general-purpose robots (XR-1 through XR-4), the low-cost Ginger Lite for reception and service, and unmanned vending robots.

Huang Xiaoqing once projected unmanned vending robot shipments in the tens of millions, and annual reception service robot volumes of at least 100,000. As of 2026, Galaxy General — a specialist in unmanned vending robotics — has not matched those deployment targets.

More troubling still, Daluo began exploring capital maneuvers almost immediately after securing investment — the goal being to build revenue and pursue a U.S. listing.

Per early employees, "Daluo built nothing substantive upfront — instead, it produced large volumes of phones, including enterprise-grade secure devices meant for Foxconn that never materialized."

From 2017 through 2018, Huang Xiaoqing wrestled with a struggling robotics business and a simultaneous Nasdaq ambition. He needed orders to satisfy investors and performance data to defend the valuation — hence the forced pivot to phone manufacturing. Mobile telecommunications, after all, was his home territory.

This chapter has been deliberately minimized, but it left an indelible imprint: Daluo's communications DNA. Huang Xiaoqing forcibly grafted China Mobile's "operator" mentality onto the robotics industry, envisioning a "China Mobile of robotics" — one that sold not hardware but "cloud brain" services, billed by data consumption.

03

Huang Xiaoqing, the Capital Player

In 2019, Daluo pursued a U.S. listing. Nasdaq was the initial target, but the NYSE made a competitive offer, waiving tens of millions in fees. Huang Xiaoqing pivoted without hesitation, targeting a $500 million raise. The prospectus failed to attract investors — but it caught the attention of the U.S. government, which placed Daluo on a sanctions list. (Leiphone)

The IPO was scuttled by entity-list sanctions. Yet the U.S. government's implicit endorsement proved a "blessing in disguise." Repatriated, Daluo commenced its domestic capital roadshow.

Shortly after the NYSE listing collapsed, the Shenzhen and Shanghai municipal governments intervened. Daluo relocated to Shanghai, securing "upwards of RMB 1 billion," 243 mu of land, and extensive preferential policies.

Several senior executives proved instrumental in this process — notably Yang Guanghua and Wang Bing, who headed Daluo's government-business (To G) operations. Yang had served alongside Huang Xiaoqing since UTStarcom; Wang was a former vice president at Kingdee Software's Beijing office. Industry peers characterized them as "masters of government-relations strategy."

This government-relations model propelled Daluo's capital roster to its zenith during 2021-2023 — but at the cost of extreme equity fragmentation: Shanghai Chengtou, Zhuhai State-owned Capital, Guangzhou Knowledge City, Ganzhou Nankang District Urban Development Group. Of Daluo's 67 identified shareholders, 57 are corporate entities.

Such equity dispersion is not unique to Daluo. The embodied intelligence firms that raised aggressively in 2024 count Tencent, BYD, Hillhouse, Sequoia, and municipal government funds among their stakeholders. The roster is longer, but the logic is identical: a "platter" structure designed to sustain continuous capital raises.

Government guidance funds dominate China's LP landscape. Daluo's trap was the absence of a "next LP" willing to step in. Today's leaders are fortunate to have "a next LP still watching." The due diligence calculus, however, is unchanged: orders, revenue, certainty.

In retrospect, the leading embodied players that raised capital early in 2024 and built momentum now face, by 2026, the same demands: orders, revenue, certainty. For many CFOs in the space, nailing down production figures and unambiguous revenue projections has become the year's primary KPI — they too have reached the stage where commercialization must be demonstrated.

04

The Zenith

Daluo peaked in 2021.

Daluo operated 12 Project Management Offices at its project peak. Huang Xiaoqing convened regular project reviews: humanoid robots, cloud brain, smart joints, digital humans, delivery bots, agricultural robots, security robots. Each project lead came equipped with a slide deck, supporting data, and a pledge of "imminent mass production."

Daluo even operated its own motor factory, enabling both cost control and customization. In 2021, Daluo cast a longer shadow than Unitree, then still emerging.

"Everyone had extensive product portfolios supporting diverse sales channels," a former PMO member recalled. He later moved to another embodied intelligence firm, where he found seven or eight PMOs, the same ritual of regular project reviews, and walls covered in status charts — red for delayed, yellow for at-risk — with ratios resembling Daluo's 2021 baseline. (Leiphone)

Daluo's six product lines were later viewed by subsequent founders as a cautionary lesson: "Don't be greedy — stay focused." In truth, focus is both a risk and a luxury.

Focus means going all-in on a single track, with both the public and investors scrutinizing that one business line. The risk: betting on the wrong horse. In embodied intelligence — a domain defined by opacity — any chosen path may lead off the trend. Moreover, insufficient diversification leaves product breadth unable to justify the valuation.

Today's leaders are likewise quietly expanding — from humanoid to quadruped, industrial to domestic, hardware to software, robots to "embodied intelligence foundation models." They frame it not as expansion but as "ecosystem building." How does this differ from Daluo's 2019 "cloud robot ecosystem" narrative?

05

Huang Xiaoqing's Foresight

The founder's vision was not lacking.

In the embodied intelligence space today, investors routinely scrutinize founders' vision. By that standard, Huang Xiaoqing was an entrepreneur of exceptional technical foresight.

During multiple 2020 addresses, Huang Xiaoqing articulated his vision for humanoid robots: "Psychologically, humans cannot accept a table that talks. Robots must share human space and human tools."

At the time, the message landed on deaf ears. In 2020, humanoid robots existed only in science fiction. Boston Dynamics' Atlas was still confined to lab backflips; Unitree had just shipped its first quadruped.

Four years later, at the 2024 World AI Conference, another embodied intelligence founder echoed the sentiment: "Humanoid is the ultimate form — only a humanoid can wield every tool humans have created." The room erupted in applause. Nearly 30 years Huang's junior, clad in the same black T-shirt, he delivered the same message.

The distinction: when Huang spoke in 2020, Daluo's humanoid existed only on blueprints. When the 2024 founder spoke, his humanoid could walk and jump — though not yet work.

Conceptually, too, Huang Xiaoqing was ahead of his time.

Beyond the cloud brain concept, Huang Xiaoqing recognized early that reducing joint costs was paramount. "The number of joints determines a robot's performance," he told media in 2020. "Joint costs must be reduced without compromising quality — only then can the robot age commence." Daluo soon established a Shanghai production and R&D base with a target of sub-RMB-1,000 joints.

By 2025, many embodied firms were just awakening to the imperative of lowering joint and dexterous-hand costs — an insight Huang Xiaoqing had captured five years prior.

Huang Xiaoqing thought too far ahead — beyond what 2020 investors could grasp, beyond what 2021 state-owned LPs would back, beyond what the 2023 Nanjing industrial fund deemed "insufficient commercialization progress." He was undone by his own prescience. Yet prescience itself is not fatal — had he endured to 2024, he would have been hailed as a visionary.

Yet, viewed differently — are today's leaders not equally "ahead of their time"?

Huang's failure to deliver stemmed from two fronts: mass production plans that repeatedly collapsed, and an equity structure that foreclosed any path to survival until realization.

06

Anatomy of a Collapse

At the summer 2023 WAIC conference, Daluo exhibited alongside a sea of demo-only booths. Its robot attempted a basketball shot before a packed audience — and missed. The crowd groaned. Intended as a display of leadership, the moment became a public embarrassment.

Thereafter, Daluo rarely surfaced in positive headlines. By early 2024, salary payments had begun to fail.

"We were deep in annual-party preparations in late January 2024, and on the 31st the salaries suddenly didn't come through," a former employee recalled. The company cited a financial system glitch. By February 8th — Lunar New Year's Eve eve — Daluo convened an emergency all-hands on Feishu. Any salary tranche above RMB 10,000 was halved. Huang Xiaoqing was absent; only two executives attended: the HR director and the CFO. Two months in, March and April salaries were suspended entirely. Hiring was frozen. In May, each employee received RMB 10,000, after which provident fund and social insurance contributions ceased.

Where was Huang Xiaoqing? By his own tally, the investment institutions he met that year "exceeded the combined total of the previous nine." His itinerary: Hefei, Tianjin, Xiamen, and as far north as Jilin. These were not cursory meetings — he convened extensively. "Whether they invest or not, I provided them a critical education on the field."

Investors, however, noted that Huang's "educational" content closely mirrored his 2019 U.S. listing pitch: cloud brain, smart joints, humanoid robots, home services. Five years on, the slide deck had been reformatted — the core logic remained unchanged.

"The only capital genuinely accessible in 2024 was government money managed by private GPs," he concluded. The subtext: private GPs were cash-poor, government money risk-averse — and Daluo was trapped squarely in the gap.

Daluo's Series C — co-led by Guangzhou Knowledge City Group and Shanghai Guosheng Investment Group — exceeded RMB 1 billion. Within six months, the capital chain fractured. According to Huang Xiaoqing, in October 2023, an institution slated to close RMB 300 million ultimately reneged.

Huang declined to name the institution, though anyone familiar with Daluo's financing history could infer: RMB 300 million is the typical ticket size for a local government industrial fund. In 2023, Daluo had planned a $500 million Hong Kong IPO — only to have approval suspended by the Nanjing Industrial Fund, a quintessential local government vehicle, citing "insufficient technology commercialization progress."

Observe the current embodied intelligence financing landscape: today's leading startups are backed by Tencent, Meituan, and Goldstone Capital. They, too, seek "capital that can tolerate uncertainty."

Yet in China, even the LP structures of industrial titans like Tencent and Meituan contain government guidance fund capital. This, therefore, is not a predicament unique to Daluo.

07

Post-Mortem

Understanding Daluo's demise requires returning to Huang Xiaoqing's CV — UTStarcom.

In the 1990s, Huang Xiaoqing co-founded UTStarcom as CTO. During the PHS era, the company — worth $10 billion with annual revenue in the tens of billions — dominated Binjiang and even challenged Huawei. Its decline, however, was rooted in internal politics.

Per UT veterans, "internal company politics" had taken hold. Zhou Shaoning (ex-Google Greater China president) and Huang Xiaoqing "had divergent philosophies." Zhou "excelled at revenue generation, R&D, and product development" — leading to "power struggles." After Zhou's departure, Huang assumed control, but "Huang operates through relationships. He has some technical knowledge, but it's not his strength."

Huang Xiaoqing transplanted this "relationship-driven" management model wholesale to Daluo.

A 2018 joiner who oversaw Daluo's middle office observed: "It's always been chaotic inside. Mini (Huang's wife and CFO) loves it when people snitch — 'so-and-so is no good.' But praise doesn't land. And they love taking credit: 'Thank goodness I...' — the credit is theirs, the problems always someone else's."

Huang's one-man rule was even more pronounced. "No matter your rank or the size of your team, on many matters you must drill down to every last detail yourself," he stated in interviews. Employees saw it differently: "Huang gets into everything — even the smallest things need his sign-off. It was exhausting." Others recalled him swearing at subordinates in meetings: "You idiots," "This is so simple..." — the language was harsh.

An investor offered a broader assessment: "Bill and Mini are decent people, but they lack business sense — both are fixated on the technology roadmap. Mini has improved somewhat in recent years. Bill hasn't."

More corrosive still was the hiring pattern. In its later stage, Daluo imported a wave of executives from Alibaba and ByteDance — "people who were impractical, addicted to formalities and upward management, flooding the system with weekly and daily reports, sometimes twice a day." Problems were not solved but escalated through layers. "Everyone knows the boss's style, so everyone buries him in granular reports. He never has quiet time to think about the company's next move."

Xie Zheng's experience encapsulated Daluo's culture. A linchpin of UBTech's Walker team, he joined Daluo in 2022 — and discovered that "the technical direction and architecture were Huang's decisions, immutable." He departed, later co-founding Yuanluo Technology with high school classmate Lian Wenzhao, who had worked at Figure AI and closed a major funding round in 2025.

A former employee categorized the workforce into four types: first, those with high tolerance whose endurance eventually broke under sustained pressure; second, the exceptionally capable — "this place doesn't deserve me, I'm done" — who walked away; third, the resigned pragmatists: "Say what you want, do what you want — but I have my boundaries"; fourth, the compliant: "I have no opinions. Whatever pleases you, I'll execute. The people below? Not my concern." By the end, the fourth type dominated.

08

Rebirth and Lessons

April 2025: a video surfaces of Huang Xiaoqing, now chairman of Gangzai Robot, being interviewed by Hong Kong media.

Now in his 60s, his hair gray, Huang recites the same script: "home services," "cloud brain," "a robot for RMB 100,000." The backdrop changed — from Shanghai's 243-mu facility to a small Hong Kong office. The title changed — from Daluo founder to Gangzai Robot chief scientist. The story did not.

Gangzai Robot is a joint venture between Daluo and Hong Kong-listed Guohua Group, which holds a controlling stake. Guohua pledged to fund the venture within one to two years — contingent on the team "building the business, technology, and products." That is the identical condition investors imposed on Huang Xiaoqing in 2015, when Daluo was newly founded.

Meanwhile, the pitch decks of today's well-funded embodied startups feature the same vocabulary: "home services," "cloud brain," "RMB 100,000 robot." The investors are the same cohort that backed Huang Xiaoqing — merely operating under different fund labels.

The difference: Huang Xiaoqing is in his 60s; today's founder, in his 30s.

Daluo's collapse was not a terminus but a cycle. The industry replays the same arc: technology ahead of its time, grand visions, funding-driven growth, government ties, product line sprawl, founder-centric control, capital chain rupture, unpaid salaries, layoffs, collapse, resurrection — and collapse again.

Today's leaders confront structural predicaments akin to Daluo's. Their good fortune: Daluo perished in 2024; they have endured to 2026. What comes after?

"Who killed the embodied intelligence unicorn?"

Not technology or roadmap: Daluo's cloud brain, smart joints, and humanoid robots were not retrograde — today's leaders tread the same ground. Not the market: the humanoid robot sector surged in 2024-2025, buoyed by genuine demand — today's leaders pitch the same narrative. Not management: Huang Xiaoqing's one-man rule was flawed, yet today's founders are near-universally "technology autocrats" bearing the same stamp.

The true culprit: the industry's underlying logic has not changed.

Ahead-of-its-time technology, grand visions, funding-driven growth, government ties: this formula propelled Daluo in 2015, unraveled it in 2019, minted new unicorns in 2024 — and will, at some future inflection point, put them to the test.

Daluo is not a "failure." It is a "trailblazer." Its mistakes are being reprised by today's firms. Its path is being retraced. Daluo's admonition to every embodied intelligence company: identify, with sober clarity, the variable that will sustain them until "realization."

That variable: Daluo never found it. The leaders of 2026 may be searching still.