The people of California deeply regret: the reliability of autonomous driving is really worrying, and the torture of unmanned taxis is so painful

On August 11, California, the most conservative region in the United States, held a hearing to discuss whether driverless taxis can legally be on the road. After six hours of debate, self-driving supporters achieved a landslide victory in a 3-1 vote

Cruise CEO Kyle Vogt (Kyle Vogt) announced in San Francisco that thousands of taxis supporting L4 autonomous driving technology will be launched in the next six months to provide all-weather services to local residents

(Source: Cruise)

However, a week after the California Department of Motor Vehicles lifted the ban on driverless taxis, it regretted its decision and immediately issued strict controls measure. Surprisingly, what have driverless taxis done in just one week to make the California Department of Motor Vehicles so angry?

In one week, Autonomous driving is driving California residents crazy

In order to convince the California government to allow driverless taxis on the road, Cruise A large amount of data was provided to prove the safety of driverless driving, but the results were...

After a week on the road in California, Cruise driverless rental There have been many accidents involving cars. Let’s take a look at a few typical cases.

First of all, at 10pm last Thursday, a driverless taxi collided with a fire truck at an intersection. Fortunately, the accident was not serious. The passengers were not seriously injured under the protection of the airbags and were sent to the hospital in time. In this regard, Cruise explained that the car had detected and identified the fire truck. And braking measures were taken, but because it was at an intersection, the accident could not be avoided

(Source: Cruise)

Okay, I think you have a reason for this, but what about the next question?

On the same day, a Cruise driverless taxi accidentally crashed into a road under construction and was firmly trapped in the cement. Since AI is not a real person after all, it cannot judge whether the cement has dried out, but the taxi company is still involved in the matter

Of course, the main responsibility should be borne by the road administration and road construction companies, because they are not set up at the road construction site Isolation facilities. Taxi companies are to blame because their maps did not capture road construction data. You know, almost all online map platforms in China can provide real-time mobile phone data. In addition to official surveying and mapping personnel, users can also upload road information

Since you have decided to develop autonomous driving, you should not ignore this point

Last Friday, Cruise had its biggest problem. Ten cars suddenly stalled at the intersection, causing the entire street to be blocked for more than 20 minutes. The explanation given by Cruise is that a music festival was being held locally, which disrupted the signal of the driverless taxi, causing the vehicle to lose data

Rewritten: This is really incredible, do cars have to stay connected all the time? ? So what should you do when going through a tunnel? Whether it is 4G base stations or 5G base stations, China has more than 60% coverage, but China cannot guarantee that signals are everywhere. In comparison, the base station coverage rate in the United States is much lower than that in China. Not only is the signal unstable, but it is also worried about interference from other signal sources. I really don’t know what Cruise is thinking about.

On August 18, the California Department of Motor Vehicles issued an order limiting the number of Cruise driverless online ride-hailing operations in the San Francisco area. No more than 50 vehicles are allowed during the day and 150 vehicles are allowed at night. At the same time, the agency also launched an investigation into Cruise and may revoke the company's driverless taxi operating qualification

Although another taxi company, Waymo, did not expose too many accidents this time, it has also experienced Many problems have occurred, such as the first fatal case of a driverless car hitting a person in 2018 and dozens of collisions with fire trucks. Rewritten content: Although the Waymo taxi company did not expose too many accidents this time, many problems have occurred before, such as the first fatal case of a driverless car hitting a person in 2018, and an accident with a fire truck. Multiple collisions

In a week, autonomous driving technology has exposed many problems, many of which are related to safety. In recent years, many countries around the world have begun to promote commercial autonomous driving technology, but in the face of so many problems, how can we use autonomous driving with confidence?

Autonomous driving is heading towards the future,There are still many difficulties to be overcome

Cars are speeding by on the road at speeds of dozens or hundreds of kilometers per hour. If you are not careful, you may cause a car accident and cause casualties. Therefore, we naturally pay more attention to safety issues

In order to achieve autonomous driving, we first need to collect data through sensors, and then we need to combine the chip with the processor and issue instructions. From a security perspective, the more sensors, the better, and the more accurate the data collected, the better. However, a large number of high-precision sensors will put greater pressure on the chip, not to mention cost issues. After Intel, Nvidia and Qualcomm entered the automotive industry, they are developing high-performance chips. Among them, the computing power of the Thor chip launched by NVIDIA has reached 2000 trillion operations per second (2000TOPS)

## (Source: NVIDIA)

In the field of radar, ultrasonic radar and millimeter wave radar have become quite common, and some vehicles priced below 150,000 yuan will be equipped with these radars. The cost of lidar is relatively high, and currently only some mid-to-high-end models are equipped with it. Domestic companies such as Huawei and Hesai are studying how to reduce the cost of lidar, hoping to allow low-end models to also have the opportunity to use lidar

Some car companies believe that radar and other sensors are redundant, and they believe that they only rely on Vision solutions are enough to achieve autonomous driving. For example, Tesla’s upcoming new Model 3 has canceled all radars. However, the complexity of the human body far exceeds that of any machine and artificial intelligence, and the imagination ability of the brain is unmatched by computers. In addition, the way we collect environmental information is not limited to the eyes, there are many organs responsible for collecting hearing, tactile and other information

(Source: Tesla) No car company uses a pure radar solution, either pure vision or fused vision of camera radar. Pure vision solutions have very high requirements for algorithms. Even autonomous driving solutions with radar cannot ensure safety, let alone pure vision algorithms. Tesla autopilot accidents are common

Some domestic companies have realized that it is currently impossible to achieve L5 level autopilot by relying solely on sensors, no matter how good the algorithm is. Therefore, these companies have launched solutions such as high-precision maps and vehicle-road collaboration. Among them, high-precision maps require continuous mapping of road data, accurate to the centimeter level, without the need for real-time collection of road information, thereby reducing the burden on the chip

Vehicle-road collaboration is to add cameras in areas with complex road conditions and traffic congestion to collect Information that cannot be detected by car cameras and radars is then sent to the car to improve driving safety

However, no matter which solution is used, an obvious problem can be seen, that is, it requires a lot of cost .

The reason why Tesla Model 3 wants to cut off all radars is because it wants to reduce costs. Vehicle-road collaboration and high-precision maps require time, manpower and material resources to be continuously improved, and it is difficult to achieve nationwide coverage in a short time.

The future of autonomous driving is bright, but it has not yet been able to completely dominate the automotive industry

Autonomous driving,

should be put back in the cageJudging from online publicity, autonomous driving seems to be very mature and can be put into commercial use immediately with just a word from the relevant department. However, this is not the case. The current commercial autonomous driving only treats some consumers as test subjects. As for whether autonomous driving should be opened for commercial use, Xiaotong believes that it should be promoted cautiously and not be too impatient. At present, many car manufacturers have launched advanced assisted driving systems, such as Huawei ADS and Xpeng XNGP, and their performance is very good

Compared to autonomous driving, which may be exaggerated, perhaps consumers today should pay more attention to high-end assisted driving functions. While autonomous driving holds great future prospects, for now it is best to limit it to the experimental stage. After all, everyone’s life is precious and should not be a guinea pig for self-driving companies and car companies. Domestic road testing of self-driving vehicles is progressing steadily, including in Beijing, Wuhan, and Guangzhou Many cities in China have begun operating self-driving online ride-hailing services. Although it can only operate in a small area at present, the price is very affordable. If you have the opportunity, you can experience it for yourself

When these car companies collect enough data and are willing to take responsibility for accidents to ensure the safety of consumers, autonomous driving technology can truly enter thousands of households

This content comes from WeChat Public account: Dianchetong (ID: dianchetong233), the author is Lost Soul Yin

Advertising statement: This article contains external jump links (including but not limited to hyperlinks, QR codes, passwords, etc. ), designed to provide more information and save screening time, the results are for reference only. Please note that all articles on this site contain this statement

The above is the detailed content of The people of California deeply regret: the reliability of autonomous driving is really worrying, and the torture of unmanned taxis is so painful. For more information, please follow other related articles on the PHP Chinese website!