Summary of Key Points
Xiaopeng, as a participant in the new national standards for power batteries and the stringent requirements for L3/L4 autonomous driving, has recently seen the implementation of what are considered the "strictest safety standards in history" (the new battery standards took effect in July, and the autonomous driving standards were officially released). He Xiaopeng believes that although the penetration rate of intelligent driving is increasing rapidly, consumers' understanding of these technologies is still unclear. These strict standards will help address the safety shortcomings of car manufacturers. Xiaopeng has been early in deploying related technologies and has impressive safety records. In the future, they plan to upgrade their AI safety system to handle extreme scenarios such as potholes on roads and road collapses, as well as release a major update to the second-generation VLA (Vehicle Localization and Awareness) system, which will enhance model capabilities, delegate L4 functions, and improve human-machine interaction.
Detailed Analysis
1. The Implementation of Two "Strictest Safety Standards in History" Raises the Bar for Automotive Safety
Since July this year, two crucial safety standards have officially come into effect or been released:
- New Power Battery Standards: Known as the strictest in history, these standards set more stringent requirements for battery resistance to collisions, high-temperature stability, and charging safety (for example, the battery pack must not catch fire immediately after a collision, and its range and safety performance must meet standards even under high-temperature conditions), directly raising the baseline for electric vehicle battery safety.
- L3/L4 Autonomous Driving Standards: The first mandatory standards in China for advanced autonomous driving, which define the safety requirements for L3 (partial automation where the driver can release control but must be ready to take over at any time) and L4 (full automation under specific conditions). These standards cover how to handle system failures and regulate human-machine interaction, preventing car manufacturers from making exaggerated claims or exposing potential technical vulnerabilities.
These two sets of standards essentially set a "red line" for the automotive industry; vehicles that do not meet them cannot be marketed. This forces car manufacturers to prioritize safety.
2. Rapid Popularization of Intelligent Driving, but Consumers Are Confused – Strict Standards Are Necessary
Data from the Ministry of Industry and Information Technology shows that the penetration rate of L2 assisted driving (semi-automated systems that require human monitoring of road conditions) has reached 70.5% (7 out of every 10 new vehicles), and NOA (Navigation On Autopilot) assistance (automatic following and lane changing on highways) has reached 34.2%. The first batch of L3-powered vehicles is also beginning to be used on roads in specific areas. However, many car owners are confused about the differences between assisted driving and autonomous driving and do not understand the capabilities of different vehicle models.
He Xiaopeng argues that these strict standards are necessary to address this issue. They require car manufacturers to not only excel in their "strong points" (such as certain intelligent functions) but also to address their "weak points" (safety in extreme scenarios) and ensure that basic safety systems are well-established. This way, consumers can rely on the standards to ensure that the vehicles they purchase are safe.
3. Why Can Xiaopeng Participate in Setting These Standards?
Xiaopeng has been researching autonomous driving and the relevant technical standards for batteries, motors, and electronic control systems (the "three electrics") for a long time. Their assisted driving systems have demonstrated low failure rates and high safety levels on highways, indicating that their technology and data are robust enough to contribute to the development of these standards.
In other words, the standards are not arbitrarily established; they involve experienced and capable companies to ensure they are both rigorous and practical, preventing car manufacturers from overreaching or consumers from being misled.
4. Xiaopeng's Future Plans: AI Safety Upgrades and VLA System Updates
Xiaopeng has revealed two key areas for improvement:
- Enhanced AI for Safety: The system will be capable of actively identifying and avoiding extreme hazards such as potholes on roads and road collapses. In the future, these scenarios may also be included in more advanced safety standards.
- Second-generation VLA 6.3.0 Update:
- More Powerful Models: The amount of parameters in the on-board model has increased by 3.5 times, making the AI system more intelligent and adaptable to different vehicle types and road conditions (e.g., recognizing foreign traffic signs).
- Integration with the Cockpit: L4-level functions from Robotaxi services will be integrated, such as starting from a stationary position, parking alongside roads, and automatically finding parking spaces in parking lots.
- Improved Human-Machine Interaction: The VLM (Vehicle Localization and Monitoring) system will use advanced models to support more intuitive commands, such as "go to the nearest coffee shop" without providing specific addresses or requesting "park by the side of the road" verbally, making operations more similar to natural language interactions.
5. The Impact of These Standards on the Industry and Consumers
For car manufacturers, these standards represent both a challenge and an opportunity. They must invest more in safety and technology; those that rely on gimmicks to sell cars may be eliminated. For consumers, buying vehicles that meet these standards means they can trust the safety features and use them with peace of mind.
Overall, the implementation of these two sets of standards marks a shift in the Chinese automotive industry from a focus on "intelligence" to a balance between intelligence and safety. As a participant, Xiaopeng is taking this opportunity to strengthen its technological advantages.
(The entire analysis is written in plain language to ensure that non-experts can easily understand the logic and implications behind the news.)