01 Automobile Huang Zehua: It is more important than redundancy to know how the system fails in actual working conditions

Communication Group |  To enter the "Sensor Group/Skateboard Chassis Group/Car Basic Software Group/Domain Controller Group" please scan the QR code at the end of the article to add Jiuzhang Assistant . Be sure to note the name of the communication group  + real name + company + position (no remarks Unable to pass friend verification)


ae9464377135c4ccea995e10290cfe8e.png

Author |  Su Qingtao

Last Wednesday, Zero One Automobile, a new energy commercial vehicle company, held a media communication meeting in Beijing.

After CEO Huang Zehua and co-founder Zhang Hongsong shared their entrepreneurial journey and some thoughts on the industry, there was a media group interview session that lasted nearly 2 hours. Originally, the author was wondering: "Would anyone be interested in asking questions about something as unsexy as a commercial vehicle? Are you interested in listening?" But what I didn't expect was that during those nearly two hours, nearly 20 media outlets participated in the interaction. Enthusiasm is very high.

There was a lot of dry stuff in that media communication meeting, and the author sorted out the parts related to unmanned driving here——

Huang Zehua, who came from an algorithm background, why did he build a car?

Huang Zehua said:

For a long time, people working on autonomous driving have focused on AI-side things such as chips and software algorithms, paying less attention to vehicles. While I was in Tucson, I worked with many OEMs, and then It was found that the biggest bottleneck for driverless commercial vehicles is not the AI ​​sector, but the vehicle.

The entire industry has spent too much money on driverless software algorithms. Now, the software has become mature, but the vehicles are still at a very basic level. For example, after driverless driving, how to digitally detect mechanical faults in vehicles? If several parts of the vehicle are replaced after damage, will it affect the operation of the autonomous driving system? These questions are still unsolved.

It can be said that the level of vehicles seriously lags behind the development of autonomous driving. These are essentially the tasks of OEMs, but traditional OEMs don’t quite understand the demand for autonomous vehicles, so it’s actually difficult to do them well. In particular, most of the existing heavy-duty trucks are mainly integrated. OEMs focus more on how to integrate the entire vehicle at low cost , and do not pay attention to product definition and software.

Why is the company called "Zero One"?

The two founders, Huang Zehua and Zhang Hongsong, had many times before from 0 to 1 and achieved the experience of being the first echelon.

Huang Zehua: During his time in Tucson, he went through several stages: researcher, technical leader, and vice president of engineering; he was responsible for perception algorithms, a full set of algorithms, software platforms, and software and hardware platforms.

Zhang Hongsong: While at Beiqi Foton Daimler, it took him 7 years to lead the company's heavy truck sales to fifth in the country; while at Sany Heavy Industry, it took only 3 years to lead the team to surpass the second-tier players.

What has the company done in the past year and a half?

Completed two heavy-duty pure electric vehicle traction platforms positioned for medium and short-distance transportation: one with wire control for driverless driving as an autonomous driving technology platform; the other without wire control for those who only need new energy vehicles company.

The most concerned data: vehicle failure conditions

Huang Zehua said:

Currently, although heavy-duty trucks are equipped with T-box and other Internet of Vehicles systems, the OEMs do not analyze and process the data sent back by these T-boxes well, and do not even master the fault data. Therefore, its after-sales problems They all rely on service stations. In this way, the efficiency of troubleshooting is very low.

In response to this problem, many people think of designing a line control redundancy. However, redundancy can only ensure that no fatal accidents occur after a failure occurs, but it cannot fundamentally avoid or reduce failures. Redundancy is very important. , but still don’t know how the vehicle system failed; what’s more, redundancy also risks causing a fight between the redundant system and the main system, thereby reducing reliability.

What we really need is not redundancy, but a statistical understanding of how vehicles fail in real-world conditions.

Zero One has a vehicle pre-diagnosis system and a predictive maintenance system on the EE of the vehicle.

In order to collect fault data more conveniently, Zero One has also installed steer-by-wire systems in the cars it provides to some non-self-driving customers without charging customers more.

Cost control ability: about 1/4 of the money spent by peers,

Achieved consistent results

Zhang Hongsong mentioned that the funds used by 01 to build the first two cars were only about 1/4 of that of the peers . Huang Zehua believes that such a large cost savings is mainly due to Zhang Hongsong's accumulation of contacts in the supply chain and experience in cost control .

Zhang Hongsong said that during the Sany period, they have integrated the cost concept into the bones of every employee. After joining Zero One, he began to train the team's cost awareness .

Zhang Hongsong gave an example:

We have a car that needs to be hauled from Wuhu to the Changsha test site for calibration, and the shipping cost is nearly 15,000. At that time, I said: "You must at least have a basic idea of ​​China's logistics freight rates. How many cents per ton per kilometer? Just 2 cents, right? Wuhu to Changsha is 700 kilometers, and a vehicle has 10 tons. The total payment is 1.5 Wan, do the math, what is your freight unit price? It’s more than 2 yuan per ton-kilometer. People who work on heavy trucks, if you can’t even figure out the freight, then why are you working on heavy trucks, right?”

Zhang Hongsong emphasized:

Cost competitiveness is always the most basic and effective competitiveness of an enterprise, and cost control must go deep into the bone marrow.

END


7277b24c6a9beb827fdee7b563eed232.png

Communication Group |   To enter "Sensor Group/Skateboard Chassis Group/Car Basic Software Group/Domain Controller Group" please scan the QR code above and add Jiuzhang Assistant . Be sure to note the name of the communication group + real name + company + position (no remarks) Unable to pass friend verification) 

write at the end

About submission

If you are interested in contributing to "Nine Chapters of Intelligent Driving" ("knowledge accumulation and compilation" type articles), please scan the QR code on the right and add the staff WeChat.

886cf599683fa7b4666ce08f32d865ef.jpeg

Note: When adding WeChat, be sure to note your real name, company, and current position. Thank you!


Quality requirements for "knowledge accumulation" manuscripts:

A: The information density is higher than most reports of most brokerage firms, and not lower than the average level of "Nine Chapters Smart Driving";

B: Information must be highly scarce, and more than 80% of the information must not be seen in other media. If it is based on public information, it must have a particularly powerful and exclusive perspective. Thank you for your understanding and support.

Recommended reading:

When "software-defined cars" encounter "hardware holding back software"

Popular Science on Autonomous Driving Synthetic Data 1: Don’t be a “disruptor” with real data, be a “leverage” 

One of the self-driving data closed-loop series: full of ideals and skinny reality

A 10,000-word long article explaining the application of large models in the field of autonomous driving

What is the difference between "removing high-precision maps" and "light high-precision maps"? What are the challenges of implementation? | One of the Nine Chapters of Autonomous Driving Essays

Engineering capability is not an insurmountable obstacle for L4 autonomous driving companies | Nine Chapters of Autonomous Driving Essay II

The "mother-in-law thinking" of OEMs forces the autonomous driving industry to deviate from the "optimal solution" | Nine chapters of autonomous driving essays 3

Guess you like

Origin blog.csdn.net/jiuzhang_0402/article/details/132551130