Mandatory National Standards for Assisted Driving Arrive: Who Will Fall Behind in Compliance Stages

The safety boundaries of advanced driver-assistance systems have long been a key unresolved issue in the industry. Recently, the Ministry of Industry and Information Technology has publicly solicited opinions on the mandatory national standard "Safety Requirements for Combined Driving Assistance Systems in Intelligent Connected Vehicles," which has instantly attracted significant attention within the industry.

The standard focuses on different functions such as single-lane, multi-lane, and pilot assistance, setting comprehensive safety technical requirements. It constructs a "three-tier safety assurance" system from three aspects: "enhancing product capability performance," "strengthening safety assurance requirements," and "regulating system usage."

Automakers, suppliers, and industry experts are actively engaged in studying and discussing this standard. Clearly, this standard will profoundly reshape the development trajectory of the driver assistance field. A relevant official from the Ministry of Industry and Information Technology emphasized that this standard fills the gap in the safety baseline for combination driver assistance system products in China, and will become a key technical basis for industry access, quality supervision, and post-event traceability, helping to comprehensively enhance the safety level of products.

How does this national standard penetrate the layers of technical haze to build a new framework for the industry that integrates a "safety baseline, competition rules, and innovation path"? After all, in the era of assisted driving, what is more important than speed is steadily driving towards the future.

　　From "vague concepts" to "precise measurements"

In the past, when discussing the safety of assisted driving, car manufacturers often preferred to use vague terms like "L2+" and "high-level assistance" to deflect attention, leaving consumers confused. Once an accident occurs, responsibility is often disputed, leading to endless arguments.

During the promotional process, some companies have misused concepts such as "high-level intelligent driving" and "zero takeover," deliberately blurring the line between "driver assistance" and "autonomous driving," downplaying the limitations of the systems. This has led to some drivers becoming complacent, resulting in dangerous behaviors such as prolonged disengagement and distraction, ultimately causing accidents and casualties.

The most groundbreaking aspect of this national standard is that it transforms the abstract concept of "safety" into quantifiable and easily testable "hard indicators." Each requirement acts like a precise yardstick, firmly targeting the technological lifeline of car manufacturers.

GaiShi Automobile found that the draft opinion first breaks down "safety" into 46 sub-items of field tests, 72-hour road tests, 8 categories of fault injection scenarios, and quantitative thresholds such as collision speeds ≤ 10 km/h. It also mandates the installation of readable DSSCDA data recording systems to provide objective data for traffic police and judicial identification, filling the domestic technical gap in functional safety for Level 2.

From the perspective of applicability, the national standard achieves "full coverage." It directly includes M category (passenger cars) and N category (commercial vehicles) under regulatory oversight, and clearly targets three types of advanced driver assistance systems: "basic single-lane, basic multi-lane, and navigation." This means that whether it is the ACC adaptive cruise control equipped in a 100,000 RMB family car or the urban NOA navigation assistance found in a 500,000 RMB luxury vehicle, both must be tested under the same rigorous safety standards.

It is worth mentioning that the national standard's stringent requirements for the "pilot combined driving assistance system" almost eliminate the practice of automakers relying on piling up functions to attract attention. After all, the scenarios covered by such systems are the most complex, and any issues can lead to dire consequences. As experts in the field of driving assistance have noted, "Currently, various automakers have uneven technical capabilities and reserves in the intelligent driving field. The introduction of this standard can effectively prevent consumers who blindly trust combined driving assistance systems from encountering severe traffic accidents."

The "countdown rule" in driver interaction is practically a tailor-made solution to address the "human-machine responsibility" issue. In the past, the steering wheel hands-off detection mechanism in some car companies had obvious flaws, taking more than ten seconds to issue an alarm, which made it easy for drivers to let their guard down.

The countdown rule for driver interaction is almost a tailor-made solution for the issue of human-machine responsibility. In the past, some models required more than ten seconds to detect hands-off the steering wheel before issuing an alert, which made it easy for drivers to let their guard down.

The national standard now clearly states: When the system detects that the hands have left the steering wheel for 5 seconds, it will issue a HOR alert. If the eyes are off the road for 10 seconds, it will upgrade the alert to an EOR prompt if the eyes have been off the road for 5 seconds within that time frame, and after 3 seconds, it will upgrade the alert again. After the upgrade, if the warning is not heeded within 5 seconds, an immediate control warning will be issued. If there are multiple violations or if the system initiates risk mitigation functions, the vehicle will be prohibited from using the combined driving assistance system for at least 30 minutes.

This progressive warning logic incorporates the principle that the driver is always the responsible entity into the code, completely eliminating the misconception that "assisted driving equals autonomous driving."

Additionally, the national standard strengthens the traceability requirements for functional safety documentation. Enterprises are required to submit a complete set of safety analysis materials covering "system description - risk assessment - safety mechanisms" and ensure bidirectional traceability of requirements. Although the standard has not yet disclosed a clause stating that "documents submitted cannot be altered," the complete traceability chain and incident data recording mechanism have already provided a technical basis for subsequent responsibility determination.

Zhu Junyu, director of the Automotive Research Department at the Advanced Manufacturing Research Center of the China Center for Information Industry Development, points out that different automakers' systems show significant disparities in reliability and stability under complex scenarios such as mixed traffic and road construction, which can easily lead users to misjudge the functional boundaries and develop an over-reliance on the systems. Additionally, "some companies misuse concepts like 'high-level intelligent driving' and 'zero takeover,' blurring the lines between 'driving assistance' and 'automatic driving,' causing drivers to lower their vigilance, resulting in dangerous behaviors such as taking their hands off the wheel or becoming distracted."

He believes that the new standards clearly establish a safety baseline for Level 2 driving assistance systems, requiring companies to specify the applicable scenarios for the systems to avoid excessive promotion that could lead to consumer misuse.

　　Who will fall behind in the compliance stage?

The implementation of national standards will bring another significant impact, fundamentally changing the "competitive logic" in the assisted driving field. In the past, the focus of industry competition was on who could be the first to launch new features, such as "the first to launch urban NOA" or "the first to achieve map-free driving." However, starting in 2027, the core of competition will shift to "who can comply first" and "who can maintain compliance."

After all, no matter how cool the features are, if they cannot pass national standard testing, they cannot be sold in the market.

The interpretation document of the national standard clearly states that the mandatory national standard for assisted driving has a defined transition period: from January 1, 2027, new type approval applications must meet most of the requirements, and from 2028, all must comply; approved models must complete rectification by 2029 at the latest.

The industry generally believes that the seemingly ample transition period is still a "test of life and death" for car manufacturers with insufficient technological reserves.

In terms of functional safety verification, some organizations estimate that if a complete team is established, documentation is supplemented, and fault injection testing is conducted, the compliance development cost for a single model can reach tens of millions of yuan. The data recording system also requires additional hardware (storage chips, backup power) and a cloud platform, which will generally increase the material cost of the entire vehicle.

This has the greatest impact on companies that have relied on "low-cost assembly" in the past, as their auxiliary driving solutions are often externally purchased and assembled, with DMS cameras and gaze tracking algorithms often being as cost-effective as possible. To meet the national standard's progressive alarm requirements for driver monitoring, hardware upgrades or algorithm retraining are necessary, both of which would erode the original price advantage.

Tesla, Huawei, XPeng, and other leading companies have already taken the lead in developing in-house DMS and functional safety systems. The implementation of national standards may actually amplify their first-mover advantage.

Therefore, industry organizations have pointed out that the phenomenon of "installing at a loss" for low-priced models under 100,000 yuan is likely to emerge first, and the concentration of the industry is expected to accelerate.

Zhang Hong, a member of the Expert Committee of the China Automobile Dealers Association, believes that in the future, L3-level models will become the focus of market competition. Subsidy policies may be tilted towards models equipped with L3-level and above advanced driving assistance systems, guiding companies to shift resources to higher-level advanced driving fields. This will promote the democratization of technology, narrow the price gap between models, and thereby accelerate the transformation of new energy vehicles from high-end consumer goods to accessible travel tools.

It is also worth noting that the national standards may reshape the trust logic of consumers. In the past, consumers obtained various information about purchasing driver assistance features primarily through the sales side. After the introduction of the national standards, consumers will only need to "base their decisions on the standards," which will prompt car companies to shift their resources from "marketing hype" to "technical compliance."

For example, a certain car manufacturer's auxiliary driving system previously caused accidents due to untimely hand-off detection. Now with the arrival of national standards, consumers can clearly determine that "this is not accidental, but due to the system not meeting standards." This shift in perception will leave non-compliant products with no place in the market.

　　How to play with new tricks under the compliance framework?

Will national standards become a "shackle" that restricts innovation? At this time, national standards are still in the stage of soliciting opinions, and such views are beginning to emerge.

Gasgoo believes that national standards are more like a "compass" guiding the direction of innovation, clearly delineating the "red lines" that must not be crossed, which in turn allows car companies to focus more on "where to go." After all, innovation without safety boundaries is essentially "running naked," and true industry leaders have always pushed innovation to the extreme within the rules.

“Compliance is undoubtedly the foundation of innovation,” said the expert in assisted driving. As for the "functional safety documentation" required by national standards, it may seem like a burden on the surface, but in reality, it can help car companies identify technical vulnerabilities. Through the pressure of standards, companies may better understand their shortcomings and make targeted improvements.

It can be expected that similar cases of "compliance driving technological improvement" will become more frequent. For example, in supplier management, national standards require that components must meet safety requirements, which will drive car manufacturers to engage in deep cooperation with sensor and chip manufacturers to jointly develop higher precision millimeter-wave radars and lower latency domain controllers, rather than casually procuring components to simply get by.

On the premise of meeting compliance requirements, the space for innovation is far broader than imagined. The key lies in identifying "user pain points." Take human-machine interaction as an example: national standards only specify "when to alarm," but do not provide specific regulations on "how to alarm," thus leaving room for innovation for car companies.

For example, I haven't seen some car manufacturers attempt to use AR-HUD to directly project "gaze return prompts" onto the windshield. Compared to traditional dashboard pop-ups, this method is more eye-catching. Additionally, some manufacturers have adopted a combination of steering wheel vibrations and voice prompts to avoid the risk of drivers ignoring a single warning method. These innovations not only comply with national standards but also significantly enhance user experience, providing far more value than simply adding features.

Another innovative direction is "data-driven security optimization." National standards require vehicles to record key data, which not only serves as an important basis for accident tracing but also as a valuable "gold mine" for optimizing systems.

In the industry, it is common to find through cloud feedback that highway tunnel entrances are high-incidence areas for driver distraction, prompting an adjustment in the EOR warning sensitivity for this scenario. By utilizing lane-changing data from fleets in different regions, the collision risk under "multi-lane, high-traffic" conditions is statistically analyzed, and the lane-changing decision algorithm is iterated accordingly.

This closed loop of "feeding algorithms with real data" has been regarded by many manufacturers and suppliers as a core method to reduce false positives and enhance comfort. The national standard establishes the "safety baseline," while the upper limit of user experience is always in the hands of companies that can fully utilize data.

Summary of the World

For car manufacturers, national standards are not the "end point," but rather a "new starting point." Only those companies that can continuously innovate within the framework of compliance can truly seize the development dividends of the intelligent driving era. For consumers, national standards serve as a "reassurance," allowing them to make informed decisions based on clear standards when choosing assisted driving features, rather than having to blindly guess in the future.

What we are looking forward to is not who introduces cooler features, but who can achieve the utmost safety. The ultimate goal of assisted driving is to make travel safer and more effortless. This national standard is undoubtedly a strong step towards achieving this goal.