Smart driving causes accidents one after another, is AEB the biggest safety hazard?

Recently, the topic of whether the AEB function is reliable has suddenly become a hot topic.

The reason why people think that the AEB function is unreliable is that it is sometimes not sensitive and sometimes too sensitive. In many recent accidents involving new energy vehicles, online public opinion has pointed the finger at AEB failure: an M7 Plus crashed into a road maintenance vehicle without AEB intervention; there was clearly no obstacle in front, but the AEB of the Ideal L9 inexplicably intervened, causing the vehicle to brake urgently and be rear-ended.

These incidents have attracted the attention of netizens. After all, driving assistance systems are now very common in new energy vehicles, and consumers are concerned about whether this seemingly high-end function is reliable. As an active safety configuration, unreliable AEB is unacceptable to everyone. So can AEB become reliable? How do car companies do it?

AEB is not a new technology, but its ability to provide accurate braking is questionable

Last year, Richard Yu and He Xiaopeng had a debate over AEB, which brought AEB, a professional term that used to be limited to the automotive industry, into the public eye. AEB technology was tried and applied in Mercedes-Benz, Volvo and other products more than a decade ago, but it was rarely as well known to the public as it is today.

The so-called AEB (Automatic Emergency Braking System) is an active safety technology for automobiles. The original intention of AEB is to actively help you brake in an emergency. If the vehicle encounters sudden danger, AEB will automatically trigger the car's braking system; if the driver has already stepped on the brakes, but if the system determines that the braking force is not enough to prevent a collision, AEB can also help increase the braking force.

In fact, countries have been promoting the popularization of AEB. As of now, about 90% of new cars sold in the United States are equipped with AEB. Although China does not require AEB as a mandatory configuration for passenger cars, the AEB installation rate of vehicles is increasing. According to the latest data from the China Passenger Car Association, the AEB function installation rate of passenger cars will reach 51.1% in 2023, an increase of 3.1% year-on-year.

However, despite its popularity, AEB does not mean that it can isolate all dangers due to the limitations of scene triggering logic. Because the triggering conditions of AEB are mainly based on the extreme calculations of distance and time, "missed recognition" and "misrecognition" may occur. It can be said that the triggering and automatic braking of the AEB system are not technical difficulties. The real challenge is when to activate the system. Too high sensitivity affects the experience; too low sensitivity may lead to missed reports and accidents.

Among them, "ghost braking" is a point that has been criticized by everyone - when the vehicle is driving at high speed, the system mistakenly thinks there is an obstacle in front of the vehicle and immediately brakes to a stop. For example, Ideal L series has experienced many such incidents, because the vehicle identified the person on the billboard as a pedestrian and triggered emergency braking. Previously, Tesla also often broke out similar accidents.

Moreover, a quick search on the Internet will reveal a large number of issues such as "AEB not activated during a car accident" and "frequent AEB triggering during daily driving is very annoying". These real cases also show that both misidentification and missed identification may cause a collision accident, so finding a balance between misidentification and missed identification is a matter that tests the technical level of the manufacturer.

We advocate advanced intelligent driving, but AEB cannot be implemented well?

If we go back a few years, the debate over AEB function would hardly have attracted such widespread attention. As the penetration rate of advanced intelligent driving increases rapidly, ordinary users' expectations for smart cars are also getting higher and higher. Although AEB and advanced intelligent driving are not the same thing, for ordinary users, the quality of AEB seems to mean the quality of intelligent driving.

As we all know, the original intention of the AEB function is to protect personal safety, but the current situation is counterproductive. Not only does it not work in all situations, it also creates more "new categories" of traffic accidents. It is reported that among the complaints about intelligent assistance systems, AEB has always been the one with the highest complaint rate. It is really unacceptable that those products named as smart electric vehicles fail even the most basic AEB.

It is not difficult to see that AEB function failure and false triggering are relatively common, and even affect the safety of users' lives, which urgently needs to be solved by car companies at the technical level. On the one hand, it is necessary to increase hardware technology research and development and continuously optimize system algorithms; on the other hand, it is necessary to increase the test dimensions and road test mileage after AEB is installed on the vehicle.

Moreover, car companies should not emphasize how powerful their autonomous driving hardware is when promoting their products. Whether they can stably and continuously provide a high-accuracy experience is the key to the reliability of high-end intelligent driving. Or we can start with the most basic AEB and make AEB really easy to use, so that everyone can trust AEB and even trust intelligent driving in the future.

For car companies, the existence of AEB is to provide a safety net for drivers, so in actual use, it is necessary to accurately identify dangers, provide timely assistance, and not cause additional trouble. Although AEB is still imperfect, with technological innovation and data accumulation, the AEB system will continue to improve, so at this stage, we still need to hold the steering wheel in our own hands at all times.