For just CNY 86,800, consumers can already purchase a vehicle equipped with LiDAR and supporting navigation assistance from parking spot to parking spot.
ADAS features are rapidly penetrating lower price segments, but does securing more orders guarantee higher profits for suppliers?
Data from GGAI Research Institute shows that in the first half of 2026, deliveries of passenger vehicles in China with pre-installed NOA (Navigation on Autopilot) reached 2.9081 million units, achieving a penetration rate of 34.10%.
Among these, deliveries of vehicles with pre-installed urban NOA (excluding software-paid options) reached 1.4180 million units, with a penetration rate of 16.63%.
While market size continues to expand, solution prices remain under sustained pressure.
For ADAS suppliers, order growth and profit growth are becoming two indicators that do not move in perfect sync.
01 Why Are ADAS Solutions Getting Cheaper?
The primary driver behind the declining price of ADAS solutions is the shift in competitive focus within the market.
In the past, urban NOA was primarily featured in mid-to-high-end models, serving as a key configuration for automakers to differentiate their products.
As technology matures, an increasing number of automakers are extending advanced driver-assistance functions into mainstream price brackets. ADAS has gradually transitioned from a selling point for a few models to a critical component in product competition.
BYD launched its "ADAS for All" strategy in 2025, promoting the integration of the God's Eye (Tianshen Zhi Yan) assistance system across models at various price points.
It should be noted that God's Eye includes different technical configurations and capability tiers, meaning not all models equipped with it possess urban NOA capabilities.
However, this trend sends a clear signal: automakers are accelerating the popularization of driver-assistance functions, and scaled application creates conditions for further optimizing system costs.
As more automakers incorporate ADAS into the configuration competition for mainstream models, procurement requirements facing suppliers have also changed.
Previously, automakers focused primarily on whether solutions could offer differentiated experiences; now, beyond functional performance, they must consider per-vehicle cost, model coverage range, and mass production delivery efficiency.
Consequently, suppliers face greater pressure on their bargaining power.
In September 2026, the China Times reported that industry sources revealed some suppliers were winning automaker orders through white-box delivery methods, charging software licensing fees at the level of one hundred yuan per vehicle.
The report also mentioned that previously, high-level driver-assistance software licensing fees generally ranged from CNY 1,000 to CNY 2,000 per vehicle.
While this data reflects pricing pressure in certain projects, it should not be interpreted as implying that all ADAS solutions in the industry have dropped to the one-hundred-yuan level.
White-box delivery varies significantly in terms of software scope, development services, hardware configuration, and subsequent maintenance responsibilities, so it cannot be simply compared to the per-vehicle price of complete mass-production solutions.
Nevertheless, the emergence of one-hundred-yuan quotes in some projects indicates intensifying competition among suppliers.
Automakers seek ADAS capabilities at lower procurement costs, while suppliers must make trade-offs between winning orders and maintaining project profitability.
02 Why Does Falling Hardware Cost Not Mean Solution Prices Cover All Costs?
The entry of ADAS solutions into lower-price vehicles is closely related to hardware cost optimization.
Continuous iteration of automotive-grade chips, steadily improving integration of domain controllers, and the scaled application of sensors such as cameras, millimeter-wave radars, and LiDARs provide space for system cost optimization.
As chip and hardware platforms upgrade, some functions can be implemented on more integrated hardware platforms, offering more room for hardware selection and cost control.
However, falling hardware costs do not mean an ADAS solution can be installed on different vehicle models without adjustment.
For suppliers, model adaptation remains ongoing engineering work. Taking perception systems as an example, sensor installation positions, heights, attitudes, and field-of-view ranges differ across vehicle models.
When camera or LiDAR installation positions or attitudes change, sensor extrinsics must be re-evaluated, and calibration and verification performed based on specific conditions to ensure accurate spatial relationships between different sensors.
Body structure, windshield glass, and external trim parts may also affect the actual working performance of sensors.
In the planning and control phase, adaptation work extends further.
Differences in vehicle dimensions, wheelbase, steering response, braking characteristics, and chassis interfaces mean that the same planning trajectory and control commands may result in different driving performances when applied to different vehicle models.
This is why the same ADAS software still requires vehicle parameter calibration, control strategy adjustments, and real-vehicle validation on different models.
If suppliers lack a unified software platform, mature model adaptation systems, and automated testing tools, adding each new vehicle project may involve substantial repetitive development and verification work.
More orders can lead to greater engineering delivery pressure.
Conversely, if perception, planning/control, and toolchains can be developed into reusable platforms, minimizing differences between models by converging them into parameter configurations and limited custom development, the marginal development cost for new models can decrease.
Therefore, what determines whether suppliers benefit from low-price competition is not just how much cheaper chips and sensors have become, but also whether software platforms are reusable, whether model adaptation is standardized, and whether test verification efficiency improves.
03 Why Doesn't Order Growth Necessarily Lead to Profit Growth?
Scaled application is indeed an important way for ADAS suppliers to reduce costs and improve profitability, but there is no inevitable relationship between scale growth and profit growth.
Momenta achieved operating revenue of CNY 1.602 billion in the first half of 2026, a year-on-year increase of 75.9%; gross profit was CNY 1.173 billion, with a gross margin of 73.2%.
Specifically, technology development service revenue reached CNY 995 million, up 81.5% year-on-year; licensing service revenue reached CNY 607 million, up 67.5% year-on-year.
Looking at the revenue structure, both technology development services and licensing services maintained growth, indicating that its business does not rely solely on a single software licensing model. However, a high gross margin does not equate to final profitability.
During the same period, Momenta's R&D expenses were CNY 1.163 billion, nearly matching the gross profit of CNY 1.173 billion; net cash outflow from operating activities was CNY 381 million.
Calculated under non-International Financial Reporting Standards (IFRS), the adjusted net loss for the period was CNY 14.097 million.
Here, another financial metric needs distinction: Momenta's net loss for the period calculated under IFRS was CNY 16.537 billion, mainly affected by fair value change losses on preferred shares and other financial liabilities, where related change losses amounted to approximately CNY 16.311 billion.
This belongs to non-cash accounting factors and cannot be directly understood as the company generating cash losses of the same magnitude from current operating activities.
After the company completed its listing, relevant preferred shares had been converted into ordinary shares.
Therefore, assessing Momenta's operational status cannot rely solely on the CNY 16.537 billion net loss under IFRS, nor solely on the CNY 14.097 million adjusted net loss.
What deserves more attention is that the company's mass-production business has formed high gross margins, but R&D and operational investments remain large, leaving operational challenges before achieving continuous stable profitability.
Horizon Robotics' situation also illustrates the investment pressure faced by ADAS enterprises.
In the first half of 2026, Horizon Robotics' continuing operations achieved operating revenue of CNY 2.055 billion, up 32.9% year-on-year; gross profit was CNY 1.356 billion, with a gross margin of 66%.
Among these, licensing and service revenue reached CNY 1.129 billion, up 52.7% year-on-year, accounting for 55% of total revenue; product and solution revenue was CNY 926 million.
However, during the same period, Horizon Robotics' R&D expenses reached CNY 2.755 billion, operating loss from continuing operations was CNY 1.672 billion, and adjusted net loss was CNY 1.671 billion.
It is necessary to distinguish that Horizon Robotics disclosed a profit for the period of CNY 3.784 billion during the same time.
This result was influenced by gains from fair value changes of financial liabilities and one-time gains arising from the deconsolidation of D-Robotics, and cannot be directly understood as main business operations having achieved profitable status.
Horizon Robotics' large scale of R&D investment reflects that ADAS chip and software platform companies still need to bear high technological inputs.
But this cannot be simply attributed to price wars, nor used to judge that all its businesses are selling at a loss.
Viewing Momenta and Horizon Robotics together reveals that ADAS suppliers have different business models; some derive main income from technology development services and software licensing, while others operate both chip products and licensing services.
Cost structures, R&D investments, and revenue recognition methods vary across different businesses.
Yet they face a common problem: as ADAS functions penetrate more vehicle models, suppliers must continuously improve the reuse rate of R&D outcomes and delivery efficiency. Otherwise, revenue growth from new orders may still struggle to cover persistently increasing investments.
04 What Kind of Suppliers Will Be Eliminated by the ADAS Price War?
Falling prices for ADAS solutions are not necessarily bad.
For automakers, lower solution costs help extend driver-assistance functions to more vehicle models; for consumers, features originally concentrated in mid-to-high-end models have the opportunity to enter more mass-market price ranges.
From an industrial development perspective, scaled application also brings more real-world road data and engineering verification opportunities.
However, price declines must be supported by cost and efficiency improvements.
If a supplier possesses a mature software platform capable of reusing core algorithms, toolchains, and testing systems across multiple vehicle projects, it may dilute R&D and delivery costs through larger installation volumes, even if per-vehicle licensing prices fall.
If a supplier requires significant custom development, repeated calibration, and real-vehicle validation for every new vehicle model connected, then low-price orders may bring not scale effects, but increased project burdens.
In reality, ADAS systems do not end upon delivery completion.
After mass production, systems require continuous problem analysis, version iterations, data backflow, and function maintenance.
If project quotes are too low without leaving sufficient room for subsequent services and long-term maintenance, suppliers may face issues such as restricted R&D investment, tight delivery resources, and declining service quality.
Therefore, future competition among ADAS suppliers will not only look at whose solution quote is lower, nor merely compare who secured more vehicle model nominations.
Whether companies can sediment technical capabilities into reusable platforms, control adaptation costs for new models, and consistently provide stable services after mass production will all influence whether enterprises ultimately obtain reasonable returns.
The fact that ADAS solutions are getting cheaper is not inherently negative.
However, when the industry continuously drives down procurement prices without simultaneously improving R&D reuse rates and engineering delivery efficiency, larger market sizes do not necessarily mean more profitable suppliers.
Price can win orders, but it does not necessarily win long-term competitiveness.
Those ultimately remaining in the market may not be the suppliers with the lowest quotes, but rather those enterprises that can truly balance technical investment, engineering efficiency, and commercial returns.