Produced by | Bullet Finance

Art Editor | Xing Jing

Reviewer | Song Wen

The competition in intelligent driving is expanding beyond screens, LiDAR, and large models into deeper-level sensors and positioning capabilities.

Whether a vehicle can automatically change lanes on the highway or complete memory parking in a parking lot is often attributed to algorithms, chips, and cameras. But in tunnels, underground parking lots, under elevated bridges, or among urban high-rises, satellite signals can be blocked at any time—and the vehicle still needs to know where it is, how fast it is moving, and what posture the body is in.

High-precision positioning and inertial measurement systems are precisely the inconspicuous yet indispensable link in this capability set.

DaoYuan Technology occupies this niche track. It provides MEMS inertial chips, IMU modules, GNSS modules, and integrated navigation systems, with customers including BYD, Li Auto, Geely, GAC, SAIC, Changan, Chery, NIO, XPeng, and other automakers, and has cumulatively delivered 4 million intelligent sensors.

This supplier—barely known to ordinary consumers—has entered the main battlefield of large-scale mass production of intelligent vehicles.

(Image / DaoYuan Technology official website)

And as the automotive supply chain shifts from "whether there are intelligent driving functions" to "whether low-cost, stable large-scale delivery is possible," the core question facing DaoYuan Technology has also changed: how can today's installation scale be converted into profit, cash flow, and bargaining power? Are the robot, construction machinery, and other "physical AI" businesses in the growth curve a new revenue curve, or do they still remain at the stage of technology capability spillover?

These are all new propositions facing DaoYuan Technology.

1. A Positioning Supplier Brought Out by Intelligent Driving

When Li Rongxi, founder of DaoYuan Technology, returned to Guangzhou to start his business, intelligent driving had not yet become the hottest label in China's auto market.

At that time, navigation and positioning was even more of a standard narrow-door business: satellite signals, inertial measurement, and attitude algorithms sounded far from ordinary consumers and usually appeared only in professional scenarios such as surveying and mapping, military industry, and drones—and finding business opportunities therein was even harder.

Let alone the combination and integration of the two. In the early days, intelligent driving was mainly concentrated in high-end models and partial functions, and high-precision positioning products were more regarded as a "bonus item."

Li Rongxi's judgment in betting on navigation and intelligent vehicles was not complicated: for a vehicle to drive without continuous human control, it must first accurately know "where it is."

Satellite navigation can provide absolute position, but after the vehicle enters tunnels, underground parking lots, or high-rise dense areas, the signal will be blocked; the inertial measurement unit estimates the vehicle's motion state through acceleration and angular velocity, supplementing the positioning link when satellite signals fail.

This seemingly inconspicuous component determines whether the intelligent driving system can continuously obtain reliable position, speed, and posture information in complex environments.

Therefore, as highway navigation, memory parking, and urban assisted driving began to enter more mass-produced models, the demand for continuous positioning, posture perception, and safety redundancy increased, and DaoYuan Technology gradually grew into an "invisible screw" in the intelligent vehicle industry chain.

Data shows that in the first three quarters of 2025, DaoYuan Technology's revenue was about 474 million yuan, a year-on-year increase of about 97.3%, already exceeding the full-year revenue in 2024; during the same period, the adjusted net loss narrowed to about 50 million yuan.

At present, DaoYuan Technology's products include MEMS inertial motion measurement chips, IMU modules, GNSS modules, and P-Box integrated navigation systems.

If its business is understood as a chain, the bottom layer is MEMS inertial motion measurement chips, the middle is IMU and GNSS modules, and the top layer is integrated navigation systems such as P-Box, as well as supporting positioning, posture, and fusion algorithms.

The chip is responsible for sensing raw motion information such as acceleration and angular velocity; the IMU packages the chip, calibration, and processing capabilities into an inertial measurement unit that can be used directly; the GNSS module provides satellite positioning; the system layer fuses inertial, satellite, wheel speed, and algorithm data to output position, speed, posture, and timing information to the vehicle.

This means that DaoYuan Technology does not sell just a certain sensor, but sells different levels of "spatiotemporal intelligence" products according to customers' development capabilities and vehicle model needs. The prospectus discloses that the company adopts a three-layer progressive architecture of chips, modules, and systems, and products can be supplied separately or delivered in combination.

In 2023 and 2024, the system business was DaoYuan Technology's main source of revenue, accounting for about 61.3% and 72.2% of total revenue respectively; the module business accounted for about 31.2% and 23.6% respectively.

At that time, DaoYuan Technology's main feature was still delivering highly integrated integrated navigation solutions to automakers, and the system business undertook more product definition, software algorithms, and engineering adaptation work.

But by the first three quarters of 2025, the company's business structure had clearly switched.

In that period, module business revenue was about 339 million yuan, a year-on-year increase of about 670%, and its revenue share rose from 18.3% in the same period of the previous year to 71.5%, becoming the company's largest source of revenue. During the same period, the company's total revenue was about 474 million yuan. This means that DaoYuan Technology's revenue growth in 2025 no longer mainly depends on system products, but comes from large-scale shipments of standardized modules.

Behind this corresponds to a change in the intelligent driving supply chain.

In the early stage of intelligent driving, automakers often needed suppliers to provide relatively complete solutions, especially when high-precision positioning had not yet been fully standardized. Integrated navigation systems could help automakers shorten development cycles and complete mass production verification of functions such as highway navigation and memory parking.

But as intelligent driving functions enter more vehicle models and procurement scale by leading automakers expands, market demand for suppliers began to shift from "customized solutions" to "standardized modules that can be replicated on a large scale with controllable costs."

The volume ramp-up of the module business has brought DaoYuan Technology's products into a larger-scale front-loading supply chain, and also means that its business model has begun to tilt from project-based delivery to manufacturing-based delivery.

This transformation has brought revenue scale, but also new operating pressure.

On the one hand, standardized modules are easier to replicate across more vehicle models, which can improve production line utilization and reduce unit manufacturing costs; but at the same time, the differentiation of module products is usually weaker than that of complete systems, and customers are more likely to negotiate around price, payment terms, and delivery conditions. Maintaining a relatively high profit margin during the module ramp-up will become the key to whether its large-scale operation can be established.

2. Major Customer Ramp-Up, Revenue Growth Has Not Yet Been Fully Realized

From the perspective of application fields, automobiles remain DaoYuan Technology's absolute base.

In 2024, the company's automotive business revenue was about 368 million yuan, accounting for about 89.8% of total revenue; non-automotive business revenue was about 42 million yuan, accounting for about 10.2%. In the first three quarters of 2025, automotive business revenue further increased to about 440 million yuan, accounting for about 92.8% of total revenue; non-automotive revenue was about 34.3 million yuan, accounting for about 7.2%.

This set of data reveals DaoYuan Technology's current true business focus: it is first and foremost an automotive intelligent components enterprise.

(Image / 123RF, based on VRF agreement (AI digital content))

The public customer list also confirms this. DaoYuan Technology's disclosed partner automakers include BYD, Li Auto, Geely, GAC, SAIC, FAW, Dongfeng, Changan, Chery, Leapmotor, NIO, XPeng, Zeekr, and AION, among others; overseas, projects have been nominated by brands including Volkswagen, BBA, Volvo, and smart.

However, broad customer coverage does not equal a "healthy" revenue structure. DaoYuan Technology's business lifeline is still pressed on a single customer.

In the first three quarters of 2025, DaoYuan Technology's top five customers contributed about 88.8% of revenue, and the largest customer contributed about 60.4% of revenue. Based on reasonable inference from the notes to the prospectus, this largest customer may be BYD.

BYD's huge shipments have given DaoYuan Technology the opportunity for rapid volume ramp-up and scale manufacturing, but have also exposed it to stronger procurement bargaining, downward price pressure, and payment term pressure.

In 2023 and 2024, DaoYuan Technology's revenue was about 366 million yuan and 410 million yuan respectively. By the first three quarters of 2025, the company's revenue rose to about 474 million yuan, a year-on-year increase of about 97.3%, with nine-month revenue already exceeding the full year of 2024.

From the perspective of narrowing losses, the company seems to be approaching economies of scale, but operating cash flow has not improved simultaneously.

In the first three quarters of 2025, DaoYuan Technology's net cash outflow from operating activities was about 320 million yuan; the company's accounts receivable were about 461 million yuan, close to the revenue scale of the same period.

This is precisely the contradiction that automotive supply chain enterprises are most likely to fall into: the larger the orders and the faster the business expansion, the higher the funding demand may be for raw materials, production, R&D, and personnel; if the customer settlement cycle is long, the company's book revenue growth will instead intensify cash occupation.

For suppliers, a fast-growing leading automaker can bring order scale, production line utilization, brand endorsement, and product verification opportunities. Especially when the intelligent driving supply chain is still in a reshuffling period, entering the mass production system of automakers such as BYD may help DaoYuan Technology lower unit costs and attract more automakers' attention.

But the problem is that when the largest customer contributes more than 60% of revenue, it is difficult for the supplier's bargaining power not to be affected. From this perspective, DaoYuan Technology's IPO will also face a growth test under customer concentration pressure.

3. Is "Physical AI" Capability Migration or a Second Growth Curve?

Beyond automobiles, DaoYuan Technology is extending its business narrative to "physical AI."

This is a concept frequently used in the past two years.

The physical AI defined by DaoYuan Technology is not a large model, but underlying sensing capability that provides intelligent machines with something like "sense of balance + sense of direction + sense of coordinates." Sensors are responsible for perceiving motion, positioning algorithms are responsible for calculating spatiotemporal coordinates, and upper-level control systems make action decisions based on these data.

In automobiles, this chain serves assisted driving and automatic parking; in robots, it can also serve balance control, motion planning, and autonomous navigation; in construction machinery, wind power, and agricultural scenarios, it can help equipment complete path planning, posture maintenance, or automatic operation.

Simply put, what DaoYuan Technology does is migrate the inertial measurement, calibration, fusion algorithms, and automotive-grade manufacturing capabilities accumulated in the automotive industry to more scenarios that require "spatiotemporal perception."

But capability migration does not equal commercialization replication.

The automotive market is characterized by longer vehicle model life cycles, large procurement volume from a single customer, and complex but relatively stable certification processes. In contrast, markets such as robots, the low-altitude economy, and intelligent machinery are still in their early stages, with rapid changes in product forms and technical routes, smaller customer procurement scale, and standards systems that are not yet mature.

DaoYuan Technology's official website has disclosed cooperation with robot companies such as Galaxy General, Galaxea, and OMOWAY, as well as projects in the field of new energy yachts.

But for the capital market, cooperation announcements are only the first step. What really matters is whether these products have obtained batch orders, whether they form sustainable revenue, whether gross margins are higher than the automotive module business, and whether customers will repurchase.

At present, automobiles will still be DaoYuan Technology's main source of revenue in the short term.

From a longer-term perspective, only after non-automotive business forms stable orders, revenue recognition, positive gross profit, and customer repurchases will “physical AI” amount to more than a technology story and become a genuine support for the company’s valuation and operating quality.

At this stage, the application version of the prospectus submitted by DaoYuan Technology in early January has expired, and the company’s subsequent capital market actions remain unknown. What is certain, however, is that the value of this company will not be determined by how many more partnerships have been signed, but by when operating cash flow turns positive, whether the revenue share of its largest customer can decline, and whether these orders can be converted into real profits.

*Note: The title image and unsigned images in this article are all from 123RF, based on the VRF agreement.