Summary of Key Issues
Recent incidents involving L2-level intelligent vehicle assisted driving in China have exposed three major problems within the industry:
1. Responsibility allocation tends to favor the car owner: Due to regulations requiring drivers to maintain full control over the vehicle at all times, automakers often use user manuals to exempt themselves from liability.
2. Significant differences in product capabilities: The cost of L2 systems varies dramatically across different models, resulting in vastly varying performance levels; consumers are essentially buying a "lottery" when choosing a car with this feature.
3. System limitations in certain scenarios: These systems struggle to recognize unusual situations, leaving room for errors in functions like Automatic Emergency Braking (AEB).
4. Consumer misconceptions: Consumers rely too heavily on assisted driving, making it difficult to prove liability in the event of an accident. The key to mitigating these risks lies in consumers adjusting their expectations and the industry implementing clear grading standards.
Why Do Car Owners Always Bear the Blame After Accidents?
The reasons are rooted in both legal regulations and the high bar for providing evidence:
- Legal Framework: All L2-level assisted driving systems sold in China are considered "assisted driving" rather than fully autonomous. Laws explicitly require drivers to keep their eyes on the road and hands on the wheel, meaning they bear primary responsibility in accidents. If authorities determine the owner was distracted or not in control of the vehicle (e.g., due to taking their hands off the wheel), they will be held accountable.
- Automakers' Exemptions: Involved companies often cite user manuals, which state that assisted driving is not autonomous and require drivers to take over at any time, thus exempting them from liability.
- Evidence Challenges: Owners facing accidents need to prove that a system failure caused the issue, but it's difficult for non-experts to provide professional evidence of such flaws. For example, in one case where a car rear-ended another, the owner claimed the system failed to warn them, while the manufacturer argued the owner was not paying enough attention; ultimately, the owner had to bear full responsibility due to lack of proof.
L2-Level Assisted Driving: A "Lottery" for Consumers?
Just because all systems are classified as L2 doesn't mean they offer similar capabilities:
- Cost Variations: The cost of L2 systems ranges from a few hundred yuan to tens of thousands of yuan, depending on the technology used (e.g., some rely on a single camera, while others use multiple cameras and lidar). Lower-cost systems have weaker algorithms and hardware, leading to poorer performance. Higher-priced systems can handle more scenarios but are not omnipotent.
- Consumer Confusion: With all cars labeled as L2-level, it's hard for ordinary consumers to distinguish between better and worse options. For instance, some systems can detect stationary construction vehicles on highways, while others cannot; some can avoid obstacles in rain, while others fail to do so—this creates a situation similar to buying a random product without knowing its quality.
- Lack of Grading Standards: There are no unified standards for L2-level systems, similar to the lack of distinctions between different driver's licenses (e.g., C1/C2). This makes it easy for consumers to mistake lower-tier systems for more advanced autonomous ones. The industry is discussing the establishment of such standards to help users make informed choices.
Weaknesses in Assisted Driving Systems
The main issue with these systems is their inability to handle unusual situations:
- What are "edge scenarios"? These are rare events that the system has not been trained to recognize, such as unexpected obstacles like tires on the road or cones in construction zones. The system's recognition ability depends on its training data; if these scenarios are not included, it will fail to detect them.
- Limitations of AEB: Although there are national standards for AEB, they allow a 10% error rate (i.e., one out of ten tests may result in a missed brake application). To avoid unnecessary braking (e.g., mistaking a plastic bag for an obstacle), manufacturers' algorithms prioritize avoiding false alarms. As a result, systems often fail to apply brakes when needed on real roads.
Tips for Consumers to Avoid Risks
To use assisted driving safely:
- Change Your Mindset: L2-level systems are not fully autonomous; always keep your hands on the wheel and eyes on the road, especially in complex situations like urban congestion or bad weather. Even if the system is allowed on highways, remain vigilant and ready to take over control at any time.
- Document Everything Before Buying: Salespeople may exaggerate the capabilities of these systems; record conversations or keep purchase receipts as evidence in case of issues.
- Know How to Seek Compensation: If AEB brakes unexpectedly (e.g., without an obvious obstacle), it may indicate a manufacturing defect, and you can report it to the relevant authorities. If it fails to apply brakes when needed, proving the defect can be challenging but possible with professional evaluation.
In summary, intelligent assisted driving is a tool, not a substitute for active driving. Always keep control of your vehicle to ensure safety. The industry needs to establish clear standards and improve technology, but in the meantime, consumers must remain cautious and use these features responsibly. After all, you are ultimately responsible for your own safety on the road.