EN / 中文

Same Model, Two AD Systems: Risks and Logic Behind Automakers’ Dual-Track Strategy

by zhijiazuiqianyan·November 17, 2026

On November 3, the Dongfeng Yipai M8 added the Tianyuan Laser Edition, featuring Dongfeng's self-developed Tianyuan intelligent driving solution. Previously, the same vehicle series had already launched a version equipped with Huawei Qiankun Intelligent Driving.

A single vehicle model offering two different sources of intelligent driving solutions. Similar product layouts can also be seen in models such as the Voyah Dreamer and GAC Trumpchi Xiangwang M8.

The former offers two major series: Kunpeng Intelligent Driving and Huawei Qiankun; the latter features both the Qiankun series powered by Huawei Qiankun Intelligent Driving and the Master Laser Radar Edition, which adopts GAC Starling Intelligent Driving ADiGO GSD driver assistance system and integrates Momenta's large model.

In the past, automakers primarily chose between self-development and supplier technologies. Now, some companies are beginning to retain both options simultaneously, allowing consumers to choose based on vehicle configuration and needs.

On the surface, this enriches product configurations. However, from an industrial perspective, it reveals that automakers are unwilling to abandon the technological advantages offered by mature suppliers, nor do they wish to fully entrust the long-term development of intelligent driving to external entities.

Is this dual-track strategy a more mature technical layout, or is it a retreat path for automakers in the competitive landscape of intelligent driving?

01. Why Are Automakers Starting to Retain Two Intelligent Driving Systems Simultaneously?

Previously, the Yipai M8 entered the market using the Huawei Qiankun Intelligent Driving solution. On November 3, Dongfeng introduced the Tianyuan Laser Edition, incorporating its self-developed Tianyuan intelligent driving system into the same vehicle series.

The two solutions are not merely differences in software versions but stem from distinct technical systems.

The Huawei Qiankun solution relies on Huawei's accumulation in perception algorithms, computing platforms, software toolchains, and system engineering. This allows automakers to leverage a mature external technical system to accelerate the deployment of advanced driver-assistance capabilities.

Tianyuan Intelligent Driving represents Dongfeng's own intelligent technology layout. Through its self-developed system, Dongfeng aims to gradually establish autonomous capabilities ranging from algorithm R&D and system development to full vehicle application.

Why not simply choose one route?

The reason is straightforward. If completely dependent on suppliers, although automakers can quickly acquire mature intelligent driving capabilities, they may be influenced by external cooperation systems regarding technology iteration, product definition, and supply chain choices.

Especially when Huawei Qiankun becomes a common technical solution adopted by multiple automotive brands, automakers must consider: if competitors can also access similar technical capabilities, what will they rely on to establish differentiation?

Suppliers can help automakers quickly acquire technical capabilities, but they may not necessarily enable automakers to form unique technical competitiveness.

Conversely, relying entirely on self-development faces challenges such as high R&D investment, long periods for technical accumulation, and complex mass production verification. Intelligent driving capabilities require extensive real-world data accumulation and continuous iteration, making it difficult to achieve a mature mass-production experience solely through short-term R&D investment.

Therefore, retaining two solutions simultaneously actually addresses two different problems.

Supplier solutions help automakers obtain mature intelligent driving capabilities faster, while self-development preserves the possibility of accumulating core technologies and adjusting product routes in the long term.

Behind this phenomenon lies an easily overlooked change: automakers previously selected only one supplier, but now they are managing multiple sets of intelligent driving technical resources simultaneously.

This is not just a change in technical routes but also an adjustment in supply chain strategies.

Automakers can select different technical solutions based on the price positioning, product lifecycle, and target users of various models, no longer needing to bind all models to the same intelligent driving system.

However, this strategy has a prerequisite: automakers must possess sufficient product definition and system integration capabilities. Otherwise, the two solutions are merely product configurations provided by different suppliers and cannot truly translate into the automaker's own technical advantages.

02. Can Running Two Intelligent Driving Systems in Parallel Reduce Costs?

From a product planning perspective, offering two intelligent driving solutions within the same vehicle series seems to cover more consumers.

Users who value mature technology and brand recognition can choose the supplier solution; those more concerned with price and configuration combinations can opt for other technical routes.

However, from an R&D and engineering standpoint, running two solutions in parallel does not mean automakers can acquire more technical choices at lower costs.

Taking the Yipai M8 as an example, although the two solutions share the same body platform and chassis foundation, the computing platforms, perception hardware, and software architectures involved in the intelligent driving systems are not identical.

The Tianyuan Laser Edition uses the Horizon Robotics J6M computing platform and perception hardware such as Hesai LiDAR; the Huawei Qiankun Edition employs Huawei's corresponding intelligent driving hardware ecosystem.

This indicates that even if two cars use the same body structure, the intelligent driving systems cannot be simply transplanted.

A complete intelligent driving system requires sensor installation and calibration, computing platform deployment, perception algorithm adaptation, vehicle control interface debugging, and full-vehicle scenario validation.

Among these, vehicle control is the most easily overlooked aspect.

After identifying the road environment, the intelligent driving system needs to generate driving trajectories and control commands, which are ultimately executed by the full-vehicle control system to handle steering, braking, and acceleration.

Different solutions have variations in planning and control algorithms. Even facing the same road environment, they may output different trajectories and control strategies.

Automakers need to verify not only whether the system functions correctly but also whether the vehicle can perform actions such as lane changing, turning, car following, and emergency avoidance as expected.

Moreover, this work does not end after the model launch.

Algorithm upgrades, sensor adjustments, and feature optimizations may involve recalibration, regression testing, and full-vehicle validation. The two systems require separate maintenance of software versions and testing frameworks, and R&D teams must allocate resources between self-development, supplier coordination, and joint validation.

Therefore, the increased cost of a dual-route strategy is not just purchasing another set of hardware or software, but the engineering investment required for development, integration, and continuous iteration of two technical solutions.

Only when the full-vehicle architecture, control interfaces, and validation standards can be unified do automakers have the opportunity to reduce repetitive development and maintenance costs.

Conversely, if each solution requires establishing a separate development and maintenance system, long-term costs will continue to rise alongside the increase in product choices.

03. What Problems Arise with Two Systems?

Compared to hardware procurement and R&D costs, the relationship between data and models is more worthy of attention, according to Leading Edge of Intelligent Driving.

Current competition in intelligent driving is no longer just about algorithm models but a systemic competition comprising data collection, scenario mining, model training, simulation validation, and mass production iteration.

Whether an intelligent driving system can continuously progress largely depends on its ability to identify problems, accumulate effective data, and convert this data into model capabilities.

However, when a single vehicle series adopts two intelligent driving systems from different sources, a question arises: can the data and technical experience generated by the two systems truly be utilized mutually?

For instance, if the Yipai M8 Qiankun Edition encounters identification issues in a complex construction scenario, can the Dongfeng Tianyuan Intelligent Driving team utilize relevant scenario data to improve their own system?

Conversely, can the processing experience accumulated by Tianyuan Intelligent Driving in certain scenarios help enhance the performance of the Qiankun solution?

This is not entirely impossible. Information such as vehicle operating status, road environment, and traffic participant behavior can, under the premises of data authorization, privacy protection, and data governance requirements, undergo unified processing to provide references for different R&D teams.

However, this depends on data authorization, usage scope, and cooperation agreements between automakers and suppliers. It cannot be assumed that because the vehicles belong to the same automaker, the two systems can automatically call each other's data.

More critically, sharing data does not imply that model capabilities can be directly transferred.

Different intelligent driving systems may adopt different sensor configurations, data formats, annotation systems, and model architectures.

The same road scenario may correspond to different perception results and decision-making problems in different systems.

Even if the same raw data is obtained, it may need to be reprocessed, annotated, and trained before being used in another system. Just because a scenario has been successfully solved by one model does not mean another model can directly acquire the same capability.

This is because models learn not only the scenario itself but also the relationship between scenario features and decision outcomes. Differences in training methods, network structures, and decision mechanisms mean that the same dataset may not produce identical learning effects.

In other words, even if the two systems can share partial data, they may still need to independently complete model training, testing, and validation.

For automakers, this is the true issue they face when operating dual routes in the long term.

If the installed scale of the self-developed solution is limited, resulting in insufficient data accumulation and scenario coverage, the speed of model iteration may be affected. While supplier solutions can provide mature driving experiences, automakers may not directly master their underlying models and training systems.

Ultimately, although a single vehicle series possesses two intelligent driving solutions, it may form two relatively independent paths of technical accumulation.

Another question is whether automakers can consolidate the experience accumulated from different models and suppliers into their own systemic capabilities.

If they cannot achieve this, possessing a large number of vehicles equipped with intelligent driving systems does not guarantee that the data generated by these vehicles will become their core technical assets.

What is truly valuable is not how much vehicle data one owns, but whether data from different sources can be transformed into sustainably reusable technical capabilities.

04. What Will Be the Final Outcome of the Dual-Route Strategy?

The product layouts of Dongfeng, Voyah, and GAC indicate that providing intelligent driving solutions from different sources within the same vehicle series is becoming another option for automakers.

However, a dual-route strategy does not necessarily signify stronger technical strength. It may ultimately lead to three different outcomes.

The first is that self-development gradually matures and begins to undertake intelligent driving development tasks for more models. As the installed scale expands, enterprises can accumulate more real-world data, gradually refining algorithms, toolchains, and full-vehicle validation systems. Once the technical capabilities and product experience of the self-developed solution reach expected levels, automakers may reduce their dependence on external suppliers.

The second is that supplier solutions maintain a long-term advantage, while self-development fails to achieve economies of scale. In this case, although automakers retain the self-development route, due to limited installed volume and continuously increasing R&D investment, they struggle to build sufficiently strong technical competitiveness. The self-developed solution may remain confined to certain models or specific configurations, turning the dual-route strategy into a configuration choice based on different product positions.

The third is the formation of long-term specialized division of labor. Automakers no longer pursue developing all algorithms and hardware themselves but instead master key capabilities such as full-vehicle architecture, system integration, product definition, and safety validation, delegating certain specialized technologies to suppliers.

This does not mean automakers lose autonomy.

Automakers do not necessarily need to develop perception algorithms, train foundational models, or manufacture intelligent driving chips themselves. However, they must at least possess the ability to define full-vehicle intelligent driving products, coordinate intelligent driving systems with chassis and other vehicle systems, conduct full-vehicle level safety validation, and manage data usage and software iteration.

Only by mastering these critical links can automakers select appropriate technical solutions based on the needs of different models without being bound by a single supplier for the long term.

These three outcomes have no absolute superiority or inferiority; the key lies in whether automakers clearly understand which capabilities they wish to master.

In the past, automakers worried about lacking sufficiently strong intelligent driving suppliers. Now, as companies like Huawei and Momenta expand their mass production scale, automakers must also face the question: as supplier technologies become increasingly mature, how much irreplaceable capability can they retain?

Leading Edge of Intelligent Driving believes that the true competitiveness of future automakers may not lie in whether they possess two intelligent driving systems, nor in whether all technologies are developed internally, but rather in whether they can grasp the initiative in system integration, data management, full-vehicle validation, and product definition.

The appearance of two intelligent driving systems in the same vehicle is superficially a change in product configuration, but behind it lies automakers' rethinking of discourse power in the intelligent driving industry chain.

The two routes can coexist, but they cannot simply operate in isolation.

If self-development remains merely a product version without forming a continuously iterative technical system; if supplier solutions are mature but automakers lack the ability to manage the entire system, then the dual-route strategy may ultimately bring not technical autonomy, but more complex R&D investments and supply chain dependencies.

Automakers may not master every technology in intelligent driving, but they must know what they have mastered and which capabilities cannot be handed over.