Author | Wu Yuqing Editor | Tian Zhe
"The case of SERES cannot be replicated; Westwell didn't have a single smart factory case before SERES," a former Westwell employee told Leiphone.
The "impossible to replicate" he mentioned points to Westwell's largest source of revenue in 2025. The 164 million yuan contributed by SERES was a one-off transaction, accounting for one-third of the company's total revenue that year.
In June this year, Westwell submitted an application to the Shenzhen Stock Exchange (SZSE) to list on the ChiNext board, aiming to become the first autonomous port vehicle stock.
The prospectus shows that from 2023 to 2025, Westwell's revenue surged from 181 million yuan to 493 million yuan, with a maximum annual growth rate of 152.34%. However, during the same period, the company's net profit attributable to shareholders accumulated losses of over 1.3 billion yuan, and the gross profit margin of its main business was 10.43%, a margin comparable to that of traditional manufacturing.
Why is a leader that has achieved "global number one" trapped in a dilemma of continuous cash bleed? Is it truly building an insurmountable moat, or has it fallen into a trap of a seemingly glamorous but actually hard-to-profit bitter business?
01The Crown of Global Number One and the Negative Ledger
The most contrasting aspect of Westwell's prospectus is the coexistence of high growth and high losses.
From 2023 to 2025, the company's operating revenues were 181 million yuan, 195 million yuan, and 493 million yuan, respectively, with a three-year compound annual growth rate (CAGR) of 64.92%, including a year-on-year surge of 152.34% in 2025.
Such a growth rate is impressive in any industry. According to a report by China Insights Consultancy (CIC), in 2025, Westwell ranked first in the industry in both revenue scale and fleet deployment scale of autonomous driving commercial vehicle solutions in global hub logistics scenarios.
The prospectus also mentions that, to date, the company's business has expanded to 30 countries and regions globally, establishing cooperative relationships with 70% of the top 20 global container ports by throughput in 2025.
However, behind the rapidly growing revenue, the company has not yet escaped the quagmire of losses. During the reporting period, the net profits attributable to shareholders were -416 million yuan, -488 million yuan, and -453 million yuan, respectively, with cumulative losses exceeding 1.3 billion yuan over the three years.
Regarding the large scale of accumulated unremunerated losses, Westwell explained: During the reporting period, operating revenue grew rapidly, but the overall scale remains small; to achieve technological iteration and improve technological reserves, the company has made substantial R&D investments; and the company's selling and administrative expenses have generally remained at a relatively high level.
Breaking it down, behind the high growth, Westwell's profitability, business model, and sustainability of growth still face multiple tests.
First is the low gross profit margin. During the reporting period, the gross profit margins of the main business were 6.14%, 10.83%, and 10.43%, respectively. Although it has improved compared to 2023, it remains at an extremely low level in the technology industry; the gross profit margin of an AI (Artificial Intelligence) company is even lower than that of many traditional manufacturing enterprises.
The core reason for the low gross profit margin lies in Westwell's business model. It is essentially a project-based company, providing customers such as ports and factories with integrated hardware and software autonomous driving solutions. The high degree of customization and the large proportion of hardware costs naturally limit the bargaining space.
The prospectus also admits that the gross profit margin is affected by factors such as project delivery cycles exceeding expectations, fluctuations in raw material prices, and changes in revenue structure. In the future, if industry competition intensifies, product delivery is delayed, or the efficiency ramp-up cycle of operational projects is long, the company may continue to face the risk of a low gross profit margin.
A former Westwell employee told Leiphone that actually, the port scenario has a closed space and controllable rules, which is conducive to scaling. "The reason for slow delivery is that on-site debugging is required, and personnel need to be stationed on-site. This time will depend on customer needs."
Customer concentration is also a significant hidden danger. From 2023 to 2025, the revenue proportion of the company's top five customers was 85.99%, 85.75%, and 81.17%, respectively, always maintaining above 80%. Among them, the revenue proportion of the largest customer was 33.93%, 48.87%, and 33.23%, respectively.
From the perspective of customer structure, Tianjin Port was the largest customer in 2023 and 2024. In 2025, SERES jumped to the first place, contributing 164 million yuan in revenue. That year, within the SERES Chongqing Super Factory, Westwell delivered a smart logistics project. This production logistics business revenue accounted for 33.23% of Westwell's total revenue in 2025.
This reality of reliance on large customers means that the company's performance will be greatly affected by the order fluctuations of a single customer. Westwell also stated in the prospectus that the company mainly provides integrated solutions to port operators, large logistics transportation enterprises, etc. A single project involves a large amount of money and a long implementation cycle, and revenue mainly comes from a few large projects. The explosive growth in revenue largely depends on the procurement rhythm of a few large customers.
Low gross profit margin combined with high customer concentration has directly led to continuous cash bleed in finances. During the reporting period, the net cash flow from operating activities was -563 million yuan, -312 million yuan, and -369 million yuan, respectively. The company explained that this is mainly because the business is in the stage of expansion and global scaling, the revenue scale is relatively low, the cost of outsourced hardware is large, and it is overlaid with continuous R&D investments. The company currently still highly relies on external financing to maintain operations.
Currently, most players in the industry are still in the investment stage, and Westwell is not the only one facing similar challenges.
R&D investment is another major reason for the losses. During the reporting period, R&D expenses were 181 million yuan, 200 million yuan, and 192 million yuan, respectively, totaling 573 million yuan. Although the proportion of R&D investment to revenue has dropped from 99.95% in 2023 to 38.98% in 2025, showing that revenue growth is gradually diluting the proportion of R&D investment, the absolute amount remains at a high level.
However, it is precisely the continuous high R&D investment that has enabled Westwell to become one of the few companies globally capable of providing integrated hardware and software autonomous commercial vehicle solutions.
In addition, the company's equity structure is also a risk point worth noting. Founder Tan Limin directly holds only 4.67% of the company's shares. By signing "Concerted Action Agreements" with six employee stock ownership platforms—Shanghai Dongjing, Shanghai Nanjing, Shanghai Bejing, Shanghai Hongjing, Shanghai Hejing, and Shanghai Zejing—he indirectly controls 22.85% of the equity, controlling a total of 27.52% of the voting rights of the company.
Westwell wrote in the prospectus: After the completion of this issuance, the control ratio of the actual controller will be further diluted, which may have an adverse impact on the stability of the company's control.
However, another former Westwell employee stated: "Tan Limin and Gaorong Capital hold the most shares. Although Mr. Tan doesn't hold many shares directly, the combined amount is quite large. He has absolute say, and he is the only actual controller."
02From Making Chips to Autonomous Driving
Almost every item in Westwell's financial structure today can find its source in its development history.
When the company was founded in 2015, Westwell's initial direction was not ports, nor even autonomous driving. A former Westwell employee recalled: "Westwell initially claimed to make chips, then did OCR (Optical Character Recognition), and then autonomous driving. It changed directions several times."
According to public reports, Westwell's initial business was neuromorphic chips. The team developed a set of deep learning algorithms and injected them into the chip to achieve "on-chip learning," that is, deep learning on the hardware side without being connected to the internet. But it is conceivable how difficult it is for a startup to compete with large companies in the chip field. At that time, the AI industry was being hyped up by the concept of facial recognition, but Westwell never found a scenario that could truly be implemented.
By a chance, Tan Limin began to explore the port direction, which also became a turning point in Westwell's destiny.
(To learn more about Westwell's entrepreneurial history, feel free to add WeChat ID Qing371_ for communication)
Westwell applied AI visual technology to the intelligent tallying of port containers, using cameras and algorithms to automatically identify container numbers, types, and damage conditions, replacing traditional manual recording.
Tan Limin once described this system: "It's like giving the port a pair of 'smart eyes', turning every container into a trackable 'smart pixel'." This system was quickly promoted and applied in dozens of ports along the eastern coast of China upon its launch.
From a more fundamental logic, Tan Limin's choice of ports was not accidental. When reviewing his original intention for entrepreneurship, he once said: "I have always believed that the essence of AI lies in liberating productivity. Only by integrating into production-restricted areas with controllable rules and rigid demand gaps can we find an entry point. This was our entry point to open the commercialization gap of port autonomous driving in 2016."
In August 2017, Westwell sparked a collaboration with ZPMC, initiating the joint R&D of autonomous driving unmanned straddle carriers. In January 2018, the world's first autonomous driving unmanned container straddle carrier was born. Westwell provided the AI solution for single-vehicle unmanned operations and the multi-vehicle dispatch system. This was Westwell's first true port autonomous driving project.
But the straddle carrier was just the beginning. In September 2018, Westwell released the world's first true full-time autonomous driving electric heavy truck, Q-Truck, in Shanghai. This vehicle adopts a pioneering cab-less whole vehicle design, with no steering wheel and no cabin. Batteries are installed at the front of the vehicle, and it can be applied to multiple scenarios such as ports, logistics parks, and mines.
From AI software to self-developed whole vehicles, Westwell completed a thrilling leap. Without a mature wire-controlled chassis, the team developed it from scratch; without a testing site, they built a test site in a garage and on coal slag ground.
Tan Limin explained this in an interview program in 2025: "The biggest difference between us and other enterprises is that we also design the vehicles ourselves. We deeply integrate the whole vehicle chassis, control system, and autonomous driving, comprehensively improving product value from dimensions such as safety and reliability." But he also emphasized: "Our goal has never been to build cars, but to reconstruct logistics efficiency with AI."
This choice established a first-mover advantage for Westwell in global ports, but it also greatly pushed up R&D investment and hardware costs.
In 2020, Q-Truck faced a real test. Six mass-produced vehicles were shipped to Laem Chabang Port in Thailand, where they achieved the world's first normalized operation of mixed autonomous and manual driving without safety officers and physical isolation. After a year of safe operation, the shore operation efficiency continued to ramp up, successfully approaching and stabilizing at the level of manual driving efficiency.
Tan Limin once reviewed: "The Q-Truck fleet has helped users become the world's first project with mixed autonomous and manual driving operations. It is currently the only scenario for normalized operation of autonomous container trucks without safety officers."
As of March 2024, the total number of containers handled by this project exceeded 476,000 TEUs (Twenty-foot Equivalent Units), and the commercial feasibility of Westwell's technology has been validated to a certain extent.
Afterwards, Westwell embarked on global expansion. From Khalifa Port in the UAE to the Port of Felixstowe in the UK, from Mexico to Chancay Port in Peru, its business covers 30 countries and regions globally, establishing cooperation with 70% of the top 20 global container ports.
In 2023, Westwell completed two rounds of financing, Series D+ and Series E, with the Series E financing amount reaching 600 million yuan, introducing strategic capital such as Ping An Capital and China Mobile Capital. Since its establishment, Westwell has undergone more than 14 rounds of financing. Also in 2023, Westwell ranked first on the list of China's Top 500 Hidden Unicorns with a valuation of 7 billion yuan.
But Westwell's IPO journey has not been smooth. It first launched A-share IPO tutoring in September 2024, but subsequently experienced three changes in tutoring agencies—from CICC to Guotai Haitong Securities, and then to SWS Securities. At the end of 2025, Westwell completed Series F+ financing and officially submitted the ChiNext prospectus in June 2026. According to a former Westwell employee, Westwell once tried to list in Hong Kong, "but was rejected in the first round of inquiries."
Tan Limin himself once divided the company's development into four stages: establishment, survival, success, and take-off, sequentially needing to solve the proof of product, proof of customers, and proof of business model. From this perspective, every transformation of Westwell is about finding a track that can prove itself.
Looking back at Westwell's development history, from chips and OCR to port autonomous driving, and from AI software to autonomous driving whole vehicles, every transformation revolves around the same goal: finding commercial scenarios where AI technology can truly create value. It is these continuous investments and strategic choices that have enabled Westwell to gradually establish its current global leading advantage in port autonomous driving, while forming distinct characteristics such as heavy R&D, project-based operations, and global delivery.
03SERES' 164 Million Yuan: A Starting Point or a Peak?
The track of port autonomous driving, on the one hand, has a double-digit annual compound growth rate, and on the other hand, a single-digit gross profit margin. It looks like a blue ocean, but only when you wade in do you know the depth is unknown.
"Doing business in ports is still okay; there is indeed demand in ports," a former Westwell employee evaluated his industry. But immediately he added: "It's just that domestic ports are too fiercely competitive."
According to the CIC report, the global market size for autonomous driving commercial vehicle solutions in hub logistics scenarios is expected to grow to 6.764 billion yuan by 2030. Other data shows that the global autonomous port operation system market will grow from 28.6 billion US dollars in 2025 to 33.3 billion US dollars in 2026. The deployment of autonomous driving trucks in Chinese ports accumulated over 2,800 units in 2025, covering 34 automated operation areas in 12 major hub ports.
Zhang Rong, President of Westwell, once mentioned in an interview: "The market we face will be at the trillion-dollar level." On the global logistics value chain, which uses containers as carriers and runs through seaports, railway hubs, land ports, airports, and factories, Westwell will provide various intelligent solutions to all participants.
From the perspective of market size, port autonomous driving is still a high-growth track. But the broad prospects of the market cannot be directly equated with the profitability prospects of the players; the track is becoming crowded.
In the first half of 2026, in addition to Westwell, multiple truck autonomous driving companies concentrated on sprinting towards the capital market. DeepDirection Technology, Zero One Automobile, and Mainline Technology successively submitted prospectuses to the Hong Kong Stock Exchange. The port autonomous driving track is moving from a steady jog to a comprehensive acceleration phase.
In terms of competitive landscape, the industry has formed two main models: Syear's intelligent driving system + self-developed components + customized chassis + vehicle-road-cloud integrated full-stack solution, and Westwell's NEV (New Energy Vehicle) + autonomous driving integrated hardware and software Q-Truck model.
The scale of competitors is generally not large. Syear, UDO, and Mainline all have teams of 100 to 200 people, while according to the aforementioned former employee, Westwell's scale has approached 800 people. Scale advantage is Westwell's moat, but it is also the source of its cost pressure.
An industry insider in the unmanned delivery industry evaluated from the perspective of product positioning: "Every closed park has its own demand characteristics, which is somewhat like a project-based system. Urban distribution autonomous vehicles choose simple, repetitive, and large-scale routes. They do not require meeting the personalized needs of every small customer, as long as they meet the most standardized needs. Therefore, the urban distribution business is easier to scale, and customer concentration is not that high."
In his view, Westwell has chosen a completely different path: "If customers are deeply bound, it is not easy for them to switch to other products. This is also a threshold. Anyone can do urban distribution autonomous vehicles now; the threshold is relatively low. But for their kind of professional customization, the market is relatively hard to enter." But he also pointed out the cost of this model: "The capacity model has certain risks. If they don't use it later, your assets will be relatively heavy."
The bigger problem is, can this business be scaled and make money in China?
In the view of the aforementioned former Westwell employee, there are two core bottlenecks hindering Westwell's scaling in China. The first is efficiency. "The efficiency of domestic docks is too high; current autonomous driving cannot match the efficiency of domestic docks."
Chinese ports are renowned globally for their operational efficiency. The container throughput and operational speed of hub ports such as Ningbo Zhoushan Port and Shanghai Port have consistently ranked among the top in the world. In such an environment, autonomous container trucks not only need to replace humans but also replace highly efficient humans.
And according to Westwell's former employee, the current operational efficiency of autonomous driving can only "reach or approach the efficiency of foreign docks," which does not yet match the high-intensity rhythm of domestic docks.
The second is cost. "Dock drivers' wages are not high enough, and there are many people working, so it cannot outcompete them. Mining can outcompete them, so mining companies have all gone public," the employee said. And replacing human labor with autonomous driving needs to be built on the premise that labor costs are high enough. In Chinese ports, the wage level of container truck drivers is far lower than that in Europe and the US, which means the cost saved by replacing manual labor with autonomous vehicles is limited, and may even fail to cover the operation and depreciation costs of the autonomous vehicles. "Westwell is operating at a loss, while the docks are not losing money. Why wouldn't they take it for free if they can?"
This logic is exactly the opposite abroad. "Drivers are very expensive abroad and relatively hard to recruit. This set of things can still be profitable at foreign docks."
Simply put, the core contradiction of port autonomous driving in the domestic market is not that the technology is inadequate, but that the economic account cannot be balanced.
The aforementioned industry insider in the unmanned delivery industry also corroborated this judgment. He compared port autonomous driving with urban distribution autonomous vehicles: "Westwell only has those customers. The customers who cannot be negotiated with are still very entangled, and we will have to talk to them sooner or later. We have enough customers. For those who think there is an efficiency problem, we might just not talk to them."
Compared with the urban last-mile unmanned delivery industry, companies like Westwell with a high degree of customization are passively serving the deep needs of a few large customers, while the latter is actively screening standardized products for a large number of customers.
In addition, according to Westwell's former employee, most of Westwell's domestic orders are operational charges, settled by the number of containers handled, paid monthly, with contracts signed for 1, 3, or 5 years. The vehicles belong to Westwell and can be held through loans from financial leasing companies. "In the early stage, the whole industry was selling cars, and later it turned into an operational model." This model means that Westwell needs to invest a large amount of capital in building and deploying vehicles first, and then slowly recover the cost through long-term operational volume.
This explains why Westwell's globalization strategy is so important. It is not just an icing on the cake; only in overseas markets can autonomous driving outcompete manual labor. But there are also unavoidable problems overseas. The aforementioned former Westwell employee pointed out: "Overseas has its own problems; they have labor unions."
The resistance of labor unions to employment substitution is an invisible barrier that any automated technology going overseas must face. The industry insider in unmanned delivery also mentioned this: "The labor union problems encountered when going overseas will be encountered by any robot. It's the same all over the world. European and American countries may be more intense, while Middle Eastern and Asian countries are slightly better, which is related to the national system and the soundness of the labor union system. Just like horse carriages being replaced by cars in the past, it takes time to adjust. If the product has value, it will always push things forward."
Another detail worth noting is Westwell's expansion into factory scenarios. In 2025, SERES contributed 164 million yuan in revenue, accounting for one-third of the company's total revenue. But the aforementioned former Westwell employee is not optimistic about this: "Before SERES, Westwell didn't have a single smart factory case."
Different from urban distribution scenarios, the migration of port autonomous driving companies to other scenarios is not a simple technological replication. Every migration requires re-understanding the production environment and re-establishing customer trust, all of which require time and cost investment. The former Westwell employee further explained: "Actually, none of the scenarios are difficult. The difficulty lies in the understanding of the production environment; autonomous driving is just a part of it."
Currently, Westwell is still continuously exploring the migration of unmanned freight technology to other scenarios such as airports. Westwell also mentioned in the prospectus that the company is one of the few enterprises that have achieved large-scale commercial landing in major logistics scenarios such as seaports, land ports, airports, and smart factories, avoiding the risk of high dependence on a single scenario and forming scale benefits of cross-scenario technological capability reuse.
However, Westwell also stated that although the company has established a leading market position in closed seaport scenarios, the large-scale commercial landing of semi-open scenarios such as airports and multimodal transport is still in the promotion stage. In 2026, Q-Truck will be put into operation in the East Operation Area of Yantian Port in Shenzhen, and Fuzhou Changle International Airport will also introduce the unmanned tractor Q-Tractor.
The door to the capital market is about to open, but the road behind it is not smooth. For Westwell, what the capital market tests is not just its global number one industry status, but whether the business model can continue to be cashed in. For the entire port autonomous driving industry, this is also a question that is still being answered.