HKUST’s Entrepreneurship and Unicorn Ecosystem (II): Flagship Ventures, the Li Zexiang Network, and Institutional Support

This article follows on from HKUST’s Entrepreneurship and Unicorn Ecosystem (I): The Entrepreneurship Center, Million-Dollar Competition, and Unicorn Day^※, taking a deeper look at the university’s most distinctive spin-offs, Prof. Li Zexiang’s entrepreneurial incubation network, and the institutional machinery — such as the Office of Knowledge Transfer and the Redbird Innovation Fund — that keeps the whole ecosystem running.

---

# 1. Which Flagship Ventures Have Earned the "Unicorn" Title?

The most recognisable unicorns in HKUST’s ecosystem are clustered in three sectors: unmanned systems, industrial AI, and machine vision. This is no coincidence: it mirrors decades of accumulated research strength in robotics, control systems, microelectronics, and computer vision within the School of Engineering, and it dovetails with the Greater Bay Area’s industrial demand, which spans everything from consumer electronics to high-end manufacturing. According to a 2022 tally by HKUST’s School of Engineering, six of the 18 companies on the Hong Kong X Foundation’s “Hong Kong Unicorns 2021” list are directly linked to HKUST engineering faculty or alumni^※. In other words, these star companies are not isolated flashes of genius but crops that grew from the same disciplinary soil. Below, we examine the flagship ventures sector by sector.

# 1.1 DJI

Founder: Frank Wang (汪滔). HKUST Connection: Wang completed both his undergraduate and master’s degrees in the Department of Electronic and Computer Engineering^※ (BEng + MPhil, ECE). As an undergraduate, he was recommended by Prof. Li Zexiang to participate in the Robocon Asia-Pacific robotics competition, which first brought him into contact with unmanned aerial vehicle research. In 2006, while still an MPhil student, Wang began building DJI, initially working out of an HKUST dormitory room^※, with his supervisor Prof. Li personally incubating and supporting the venture. After incorporating in Shenzhen, DJI’s deep expertise in rotorcraft motion control and robotic perception quickly established it as the industry leader.

Major Achievements: It holds over 70% of the global consumer drone market^※, employs over 5,000 people worldwide^※; in 2014, Frank Wang was named by Forbes as one of “China’s Top Ten Innovators”^※; in 2019, Wang and Prof. Li were jointly awarded the IEEE Robotics and Automation Award^※, recognising their contribution to the commercialisation of civilian drones and aerial imaging technology. DJI also funds several HKUST postgraduate robotics scholarships^※ as a way of giving back to its alma mater.

Recent Scale (figures current to 2025): As the largest venture spawned by the HKUST ecosystem, DJI has expanded from a single-product drone maker into handheld imaging (Osmo), agricultural spraying, and enterprise solutions. According to third-party estimates, DJI’s 2024 revenue was around US$3.5 billion with a global workforce of over 10,000^※; it holds roughly 60% of the global drone market and about 80% of the US consumer market, with an even higher share in specific consumer segments^※. It should be noted that DJI is privately held and does not publicly disclose audited financial statements; the above figures are estimates from third-party research firms and should be taken as indicative rather than official. From a rotor experiment in an HKUST dorm room to a global champion that has perfected the “Shenzhen manufacturing + Greater Bay Area supply chain” model, DJI is often held up as the quintessential example of hard-tech entrepreneurship from the “HKUST camp.”

# 1.2 Googol Technology (固高科技)

Co-founders: Prof. Li Zexiang and Prof. Ko Ping-Keung (高秉强), both faculty members in HKUST’s Department of Electronic and Computer Engineering. Founded: 1999^※ (some early-phase reports cite a precursor launched in 1997, with the company formally joining HKUST’s entrepreneurship programme in 1999). Positioning: The first high-tech enterprise in the Asia-Pacific region focused on motion controllers and controller-based systems^※. Listing Milestone: On 15 August 2023, Googol Technology listed on the ChiNext board of the Shenzhen Stock Exchange^※, achieving an IPO roughly 24 years after its founding and completing its exit. It was the first independently listed company from the “Li Zexiang stable.” According to financial media reports, on its listing day Googol’s issue price was RMB 12, it opened roughly 525% higher, and its market capitalisation briefly reached around RMB 28 billion^※ (stock code 301510).

Business Foundation: Googol has long focused on motion controllers, the “invisible joints” of manufacturing. Media reports indicate that by the time of its listing, the company had deployed over 600,000 motion control systems across nearly 60 industries and more than 2,000 clients, with revenue growing from RMB 283 million to RMB 348 million between 2020 and 2022^※. As one of the technical sources through which Frank Wang first encountered robotic control as a student, Googol and DJI emerged from the same tutelage under Prof. Li, forming an early example of a “single mentor, two distinct tracks” model.

# 1.3 Yunzhou Tech (云洲智能)

Founder: ZHANG Yunfei (张云飞). HKUST Connection: Zhang came to HKUST in 2007^※ and completed an MPhil in the Department of Mechanical Engineering in 2009 and a PhD in the same department in 2020. He began his unmanned surface vessel research during his doctoral studies and, with the support of fellow students, formally founded China’s first unmanned boat company in 2010^※.

Major Achievements: Won the championship in the 2013 National Innovation and Entrepreneurship Competition^※, China’s top-tier research competition; in 2014, was selected for the “World Youth Innovation Top 100”^※; has expanded operations to over 30 countries and regions^※, serving clients in scientific research, environmental protection, hydrology, and nuclear radiation monitoring; participated in Antarctic scientific expeditions (aboard the Xue Long icebreaker) and emergency responses including the Tianjin port explosion and water-pollution incidents in Gansu and Henan provinces; has reached unicorn valuation and is developing an “intelligent mobile marine stereo observation system” integrating drones, unmanned boats, and submersibles.

Recent Developments (current to 2025): Yunzhou has pushed unmanned boats from research tools into broader commercial and consumer scenarios. Media reports state that its product range has been sold to more than 60 countries and regions, with roughly 400 authorised patents^※; in 2024 it launched the “Bubble Dolphin” smart leisure boat for tourism, deploying batches of them at scenic spots such as Wuhan’s East Lake. On the manufacturing front, in April 2025 Yunzhou broke ground on a joint medium-to-large unmanned boat manufacturing base in Hefei with a planned annual capacity of nearly a thousand vessels^※, described as China’s first large-scale production base for unmanned boats. Zhang Yunfei’s journey from MPhil to PhD spanned more than a decade in HKUST’s Department of Mechanical Engineering, making Yunzhou one of the rare unmanned-systems enterprises in the “HKUST camp” built from scratch to industry leader by a single founder.

# 1.4 SmartMore

Founded in @[December 2019 by Prof. Jiaya Jia (贾佳亚), now Dean of HKUST’s von Neumann Institute (see Robotics and Artificial Intelligence^※), with its global headquarters in the Hong Kong Science Park](https://www.hkic.org.hk/zh-Hans/latest-updates/hkic_smartMore), SmartMore focuses on machine vision and AI-driven quality inspection for industrial manufacturing^※. The company reached unicorn valuation in under 18 months, and on 12 June 2024 signed a strategic cooperation agreement with the Hong Kong Investment Corporation Limited (HKIC, a wholly-owned entity of the Hong Kong SAR government), becoming the first company HKIC invested in after its establishment^※. The signing ceremony was witnessed by Financial Secretary Paul Chan, who remarked in his speech that the partnership between HKIC and SmartMore “is a very good start.”

Technical Positioning and Clients: SmartMore builds industrial AI agents using an integrated full-stack capability spanning optics, mechanics, electronics, computing, and software, bringing large-model capabilities to industrial “blind spots” like quality inspection and defect detection. According to its prospectus materials and financial media, its client list includes leading manufacturers such as Apple, Tesla, Carl Zeiss, BYD, and Luxshare Precision^※.

IPO Sprint (2026): Financial media report that SmartMore filed its listing application with the Hong Kong Stock Exchange in March 2026, aiming for a main-board IPO as the “first industrial AI agent stock,” with a post-latest-round valuation of about US$1.23 billion^※, jointly sponsored by Morgan Stanley, CICC, and Deutsche Bank. Its prospectus also discloses that revenue grew from RMB 485 million in 2023 to RMB 1.086 billion in 2025, though it remains in a phase of widening losses^※ — a scenario that mirrors the universal tension between scaling up and reaching profitability for deep-tech unicorns.

# 1.5 APT Electronics and Other Companies on the Unicorns HK 2021 List

APT Electronics (APT) was showcased as a representative company at HKUST’s first Unicorn Day in 2023^※ and is an HKUST-linked enterprise in the intelligent vision hardware space, specialising in high-end industrial vision products. Together with SmartMore and SmartSens, it forms part of a multi-node layout that spans an entire industrial chain segment from machine vision to image sensing to AI-driven quality inspection: from CMOS sensors at the bottom (SmartSens), to vision hardware (APT), to quality-inspection AI agents (SmartMore). This neatly strings together a near-complete vision-industry chain, and indirectly validates the observation that a university’s disciplinary base defines which sectors its spin-offs cluster in.

According to a 2022 report from HKUST’s School of Engineering^※, six of the 18 companies on the Hong Kong X Foundation’s “Hong Kong Unicorns 2021” list are directly linked to HKUST engineering faculty or alumni:

Additionally, Shenzhen Unity Drive Innovation Technology (UDI) was co-founded by Prof. Liu Ming of HKUST(GZ) and Assistant Prof. Wang Lujia of HKUST (Clear Water Bay). The company won the “Best Innovation and Technology Company in the Greater Bay Area” award^※, taking the stage alongside Huawei, Xiaomi, and SenseTime.

# 1.6 Hai Robotics (海柔创新)

Incubation Background: Hai Robotics is the unicorn that emerged from the “Li Zexiang network” in the warehousing and logistics sector. According to media reports, the company was founded in 2016 in Dongguan’s Songshan Lake, with founder Chen Yuqi starting out under Prof. Li’s wing and Li serving as chief advisor^※. Hai Robotics entered the warehouse-automation market with an “autonomous case-handling robot” (ACR) solution, seen as a new path distinct from traditional AGVs or goods-to-person (GTP) systems.

Valuation and Standing: Media reports state that Hai Robotics has raised over US$300 million cumulatively, with a post-money valuation after its 2022 Series D+ round of approximately US$2.08 billion^※; its own website prominently flags a growth rate that saw it reach a RMB 10 billion valuation within five years of founding^※ and it was listed on the Hurun Global Unicorn Index 2021. Hai Robotics’ significance lies in proving that the “Li Zexiang network’s” incubation capacity is not limited to drones and industrial controls — it can also be replicated in a completely new subsector like warehouse robotics.

---

# 2. How Did Li Zexiang’s Entrepreneurial Incubation Network Take Shape?

Prof. Li Zexiang is an unavoidable nexus in HKUST’s unicorn ecosystem. According to HKUST reporting^※, the unicorns he has directly co-founded or incubated include DJI, Googol, CiDi, Hai Robotics, and 4Paradigm. One individual’s involvement in five unicorns is a rarity in academic entrepreneurship globally.

His core methodology might be summed up as “research as entrepreneurship”: using HKUST’s research base in robotics and electronic control as the technical bedrock, and leveraging the manufacturing supply chain of the Greater Bay Area to push laboratory results rapidly towards commercialisation. This “Li Zexiang model” incubation pathway has become a methodological template for HKUST’s entrepreneurial ecosystem, providing a replicable formula for later founders such as Jiaya Jia with SmartMore. Prof. Li has accordingly been dubbed China’s “hidden godfather of entrepreneurship” (for his educational philosophy on robotics, see Robotics and Artificial Intelligence^※).

# 2.1 From “Supervisor-and-Student” to Batch Incubation: QKM, AgileX, and More

The “Li Zexiang network” is not just a handful of star companies, but a web of startups woven from HKUST’s supervisor-student relationships. Frank Wang’s fellow student Shi Jinbo, also under Prof. Li’s tutelage, founded the industrial robotics company QKM Automation (李群自动化) in 2011, with a product line covering parallel, SCARA, and six-axis robots^※. A cluster of other hard-tech firms like AgileX Robotics (松灵机器人) and ePropulsion (逸动科技) also mostly grew along the same trajectory: lab research topic → Greater Bay Area supply chain → niche-market leader. This chain of “supervisor + students” incubation continuously converts HKUST’s robotics research manpower into entrepreneurial manpower.

# 2.2 XbotPark: Moving Incubation off Campus and into the Industrial Belt

Prof. Li has further institutionalised and scaled up the incubation experience validated at HKUST. According to official records and media reports, in 2014 he launched the XbotPark robotics industrial base in Dongguan’s Songshan Lake, offering hard-tech startups one-stop support from supply chain and engineering to funding^※. Media estimates suggest that in its early phase (2014–2020), XbotPark incubated over 60 teams with a survival rate of roughly 80% and a unicorn or near-unicorn ratio of about 15%^※; the Li Zexiang team has cumulatively participated in incubating over 270 hard-tech enterprises with an aggregate valuation exceeding RMB 350 billion, including several listed companies and more than ten unicorns or near-unicorns^※ (media estimates; indicative order-of-magnitude, exact figures vary by source). He later launched the Shenzhen InnoX Academy, extending the same model of engineering-entrepreneurship education — “build the product first, discuss the thesis later.”

For HKUST, XbotPark’s significance is that it upgraded a sporadic story — “one dorm room, one company” — into a replicable industrialisation process. This is what distinguishes HKUST’s entrepreneurship ecosystem from simply “having produced a few star alumni.”

---

# 3. From Unicorn Status to Listing: The HKUST Camp’s “Exit Wave”

In Part I, “22 exits” was an abstract statistic. When mapped to specific companies, it becomes a string of IPOs across the Shanghai, Shenzhen, and Hong Kong exchanges. Unicorn valuation is a paper milestone; what truly completes the “capital loop” is a public listing or a trade-sale exit. In recent years, HKUST-linked listing cases have clustered in the 2022–2026 window, forming a verifiable exit wave.

SmartSens: Founded by Xu Chen (ECE MPhil & PhD), specialising in CMOS image sensors. According to media, SmartSens listed on the Shanghai Stock Exchange’s STAR Market in May 2022 (stock code 688213); it is a leading supplier in the security CIS sector and in 2024 achieved a monthly shipment volume exceeding 100 million image-sensor chips^※.

Googol Technology: See section 1.2 above; listed on ChiNext in August 2023, it was the first independent IPO from the “Li Zexiang stable.”^※

4Paradigm: Founded by HKUST CSE PhD alumnus Dai Wenyuan, with deep involvement from HKUST Prof. Yang Qiang; it specialises in enterprise decision-making AI. Financial media report that 4Paradigm listed on the main board of the Hong Kong Stock Exchange on 28 September 2023 (stock code 06682), with an IPO price of HK$55.60 per share, raising around HK$1 billion^※, and was billed as the “first decision-making AI stock.” Third-party research indicates that in 2022, 4Paradigm held roughly 22.6% of China’s market for platform-centric decision-making AI, ranking first; its core “Prophet” AI platform generated roughly RMB 2.51 billion in revenue in 2023, up about 68% year-on-year^※. Dai Wenyuan was named a “35 Under 35 Innovator” by MIT Technology Review, while Yang Qiang is an internationally recognised authority in federated learning and transfer learning (for his academic background, see Robotics and Artificial Intelligence^※).

SmartMore: See section 1.4 above; filed its listing application with the Hong Kong Stock Exchange in March 2026, aiming to become the “first industrial AI agent stock.”^※ The application is still under review.

What this table illustrates is not simply that “HKUST produced a few good companies” — it shows that the ecosystem has successfully run the complete capital chain from lab bench and unicorn valuation through to the public market, with exit venues spanning Shanghai, Shenzhen, and Hong Kong. This is also why Part I was able to report a year-by-year increase in the “number of exits.”

---

# 4. What Institutional Machinery Supports This Ecosystem?

HKUST’s entrepreneurship ecosystem is not sustained by the sporadic genius of star founders alone, but rests on a set of systemic support mechanisms:

The Office of Knowledge Transfer (OKT) is responsible for coordinating patenting, technology licensing, and startup incubation, and serves as the administrative hub for the entire machinery. HKUST’s IP utilisation rate exceeds 30% (based on a five-year rolling window, as of 2026)^※, comparable to the world’s top research universities; over five years it filed roughly 1,670 patent applications and secured about 700 grants^※. Industry-collaborative research accounts for 6.2% of activity, more than double the global average of 2.7%^※. In practical terms, the OKT coordinates patent filing and maintenance, technology licensing, spin-off company formation, and startup incubation^※, turning scattered lab research into assets that can be licensed, capitalised, and built into companies. This “technology transfer office” model is a standard institutional pillar that top Western research universities use to convert papers into companies; HKUST has localised it and connected it directly to the industrial demand of the Greater Bay Area.

The Redbird Innovation Fund is a venture fund set up specifically by HKUST for high-potential deep-tech startups, with a scale of HK$500 million^※. It acts as a “systemic amplifier” for deep technology, providing direct capital support from seed to growth stage. In addition, the university is actively planning “InnoBay” and an on-campus DeepTech incubator.

The government’s Technology Start-up Support Scheme for Universities (TSSSU) is a vital supplementary source of external funding. HKUST has been the largest beneficiary institution of TSSSU since its launch in 2014, with associated startups receiving a cumulative total of roughly HK$400 million in grants^※.

The Million-Dollar Entrepreneurship Competition and Unicorn Day serve as the funnel’s entry point and the annual review platform for early-stage ventures; see HKUST’s Entrepreneurship and Unicorn Ecosystem (I)^※ for details.

The Greater Bay Area Manufacturing Chain — the most easily overlooked layer of support in the ecosystem. HKUST sits on Clear Water Bay, separated by a river from the electronics manufacturing belt of Shenzhen and Dongguan. This geographical chain — from basic research in Hong Kong, to prototyping and engineering in Shenzhen, to mass production in Dongguan — is the physical substrate that allows many HKUST-linked hard-tech firms to “build a prototype in three months and reach the market in six.” Prof. Li has institutionalised this advantage: according to media, the robotics industrial base he built in Dongguan’s Songshan Lake provides startups with “one-stop service and full-spectrum support,” from supply-chain matchmaking to engineering validation^※. Viewed from another angle, the reason HKUST’s entrepreneurial ecosystem can repeatedly produce hardware unicorns isn’t just on-campus patents and capital — it is also this off-campus industrial chain, a resource that is difficult to replicate elsewhere and one that purely software-focused incubation models lack.

---

# 5. How Does This Scorecard Compare with Peer Universities?

Among research universities in mainland China, Hong Kong, and Macau, having a record of unicorn incubation is not uncommon, but HKUST’s special feature is its relative youth: it was founded only in 1991. In 2022, HKUST ranked second in a Chinese university ranking of “deep-tech unicorn incubation capacity” by Xin Cai magazine^※. In terms of ecosystem density, with over 1,900 active startups, more than 11 unicorns, and 22 exits from a full-time student body of under 20,000, the number of unicorns incubated per thousand enrolled students places HKUST in the top tier among Asia-Pacific universities. Meanwhile, Unicorn Day, as a formalised annual platform, provides systematic public data disclosure that makes this quantitative scorecard relatively traceable and comparable.

A caveat: cross-university comparisons should treat differences in counting methodology with care. HKUST’s figure of “over 1,900 active startups and 11 unicorns” uses the broader boundary of “founded by HKUST members (faculty/students/alumni),” which is not directly comparable to narrower definitions used by some universities that only count a company if it was registered while the founder was still a current student. “Unicorn” valuations themselves also fluctuate with the temperature of the primary market. Rather than fixating on absolute rankings, therefore, a more persuasive observation lies in the trend and the structure: a university just 35 years old, with fewer than 20,000 full-time students, has simultaneously produced segment leaders across three hard-tech tracks — unmanned systems, industrial AI, and machine vision — and has shepherded several of them onto the public markets in Shanghai, Shenzhen, and Hong Kong. It is this density of layered factors — disciplinary base + industrial-geography advantage + institutional scaffolding — that represents the hardest part of HKUST’s scorecard to replicate.

---

# 6. What Real-World Challenges Does This Ecosystem Face?

Alongside an impressive scorecard, there are several tensions within HKUST’s entrepreneurship ecosystem that deserve an honest assessment, if only to avoid depicting the “unicorn factory” as a one-dimensional success story.

First, deep tech faces a long-running gap between “scale” and “profit.” A unicorn valuation is the primary market’s pricing of future potential, not a confirmation of current profits. SmartMore is a case in point: its prospectus shows that while revenue doubled between 2023 and 2025, net losses continued to widen over the same period, with a cumulative three-year net loss exceeding RMB 2.2 billion^※. In sectors like industrial AI and robotics, which are R&D-heavy and require long gestation, a trade-off between “scale first” and “becoming profitable” is almost universal; HKUST-linked firms are no exception.

Second, the incubation network is highly dependent on a single pivotal figure. The dense output from the “Li Zexiang network” is simultaneously a strength and a vulnerability. Whether an incubation model that relies so heavily on one individual mentor’s personal appeal, connections, and judgement can, in his absence, be stably replicated through institutionalised processes remains an open question to be tested over time. Ventures like XbotPark and the Shenzhen InnoX Academy are, in a sense, attempts to translate the experience of the individual “Li Zexiang” into a transferable organisational capability.

Third, the scrutiny of the public markets has only just begun. Unicorn valuations must ultimately face the recalibration of public-market pricing. Googol’s stock may have surged on its first trading day, but post-listing valuation corrections and results that fail to match expectations are hardly rare for hard-tech firms. Whether 4Paradigm and SmartMore can sustainably deliver growth after listing is the next test of whether this ecosystem’s “capital loop” has truly been validated.

Laying out these three points side by side is not a negation of the achievements, but an attempt to make the narrative of “from campus project to unicorn” more three-dimensional. This chain has been repeatedly proven effective — yet every link in it, from profitability and succession to exit, continues to be tested in real time.

---

# 7. Summing Up: A Complete Chain from “Campus Project” to “Unicorn”

Reading Parts I and II together reveals HKUST’s entrepreneurial ecosystem as a complete chain: The Entrepreneurship Center (founded 1999) cultivates entrepreneurial awareness and foundational skills; the Million-Dollar Entrepreneurship Competition is the funnel entry for early-stage projects; the incubation networks of Prof. Li Zexiang and others, together with XbotPark, convert lab results into companies at scale; the Office of Knowledge Transfer, Redbird Innovation Fund, and TSSSU provide systemic support in patents, capital, and government-matched resources; the IPOs of Googol, 4Paradigm, and SmartSens close the capital loop from unicorn valuation to public market; and Unicorn Day serves as the annual, platformised, and quantifiable stage on which to present all these fruits to the outside world. DJI, Googol, Yunzhou, SmartMore, Hai Robotics, and the other firms are the fruits borne by this chain after more than thirty years of continuous operation.

---

# Related Reading

• HKUST’s Entrepreneurship and Unicorn Ecosystem (I): The Entrepreneurship Center, Million-Dollar Competition, and Unicorn Day^※: The companion piece to this article, covering the “entry layer” of entrepreneurship and the quantitative macro-picture.
• Robotics and Artificial Intelligence^※: Prof. Li Zexiang’s educational philosophy on robotics, and the academic background of Prof. Jiaya Jia and the von Neumann Institute.
• Historical Timeline^※: The temporal coordinates for key milestones like the 1999 “founding of the Entrepreneurship Center.”
• AEARU and Regional Alliances^※: HKUST’s institutional logic of “expanding concentric circles” in internationalisation, mirrored by the “inside-out” trajectory of its entrepreneurship competitions.

---