Xiao Ma Zhika is coming! China’s valuation automatic driving unicorn first exposure truck strategy

China has run the fastest automatic unicorn – Xiao Ma Zhi, published truck strategy in this week.

According to the latest data, Xiaoma Zhi’s cumulative test mileage has exceeded 5 million kilometers, the company is estimated to exceed 5.3 billion US dollars, and the accumulated financing is more than $ 1.1 billion.

In the truck automatic driving field, Xiaoma Zhi’s exploration begins in August 2018.

In December 2020, the truck division department was formally established interior.

And on March 31, March 31, Xiaoma Zhi’s truck business to the official name: Xiao Ma Zhika Ponytron.

The truck business department is responsible by the vice president of Xiao Ma Zhi, is responsible. At present, the small Malaysian team has exceeded 500 people, which are nearly 50 teams responsible for truck business.

Xiao Ma Zhika works closely with OEM and Tier 1 in R & D to create an automatic driving truck in front mass production.

In commercialization, Pony believes that the smart logistics network is established through the automatic driving truck will greatly improve the cost structure and safety of freight.

According to the schedule of the official announcement, Xiao Ma Zhiqi should realize the front mass production of the automatic driving truck in about 3 years, to 2025, to achieve the production of automatic driving truck mass production.

1, Xiao Ma Zhika: Technical first, not in a hurry

At the communication meeting, the Pony first showcase the video clip of the automatic driving truck in China.

Xiao Ma Zhika tested in China

Xiao Ma Zhika tested in the United States

In the video, on the domestic road, the small horse automatic driving truck is accurately bypass pedestrians who stopped next to the road when they have been curved.

On the United States, the small horse automatic driving truck accurately identifies the ramp, and travels in a parallel traffic, and travels smoothly at 50 miles per hour (equivalent to 80 km / h) per hour.

First look at the sensor configuration on the small horse automatic driving truck prototype.

2 laser radar, currently using the 64-wire laser radar supplied by Haiyi Technology;

6 cameras, including 2 mid-range cameras, 2 depth visual long distance cameras, 2 backward cameras;

3 millimeter-wave radar, including 1 forward millimeter wave radar, 2 rear mm radar;

2 GNSS antennas.

Their role is:

2 laser radar covers object detection and ranging from the body of 360 ° 150 meters;

6 cameras, responsible for detection from far to nearly near, for supporting vehicles driving in high-speed intervals, helping vehicle changes. Two depth visual long distance cameras cover the range of 300 meters to 1000 meters; 2 mid-range cameras covers 200 meters range; 2 rear cameras over 3 meters from 200 meters.

3 millimeter-wave radar, one forward millimeter wavele covers 200 meters, 2 rear to millimeter radar cover 200 meters, and then to millimeter waves can provide assistive effects when making a variation.

Two GNSS antennas are used in cm-level precision positioning.

In the above sensors, the pony sorted a deep visual long distance camera and the double-eyed visual algorithm behind.

It can be seen that the Software and hardware system of the small Malaxing automatic driving truck is a pulse.

2, why do you do a truck? You can reuse 80% of the technology of Robotaxi, automatic driving commercialization will accelerate

Pony believes that from technical perspective, 80% of the technology in automatic driving of passenger cars can be multiplexed on trucks, but because of the operational characteristics of trucks in terms of load, with hanging and other aspects of control accuracy requirements High, so some changes are made on the algorithm.

From the end of 2016 to now, the Pony has run more than 5 million kilometers. The data disclosed in the middle of March in March is more than 4.5 million kilometers.

Rough estimation, the Pony ran 500,000 kilometers in one month in March, and this data is quite considerable.

The rapid growth of road test mileage may be related to the large growth of the team.

In early February this year, Xiaoma announced the progress of the standardized production line. At that time, a group of automatic driving vehicles equipped with its new generation of ALPHAX and opened public test in Guangzhou, Beijing, Shanghai, the number of new teams at the time. It may be nearly a hundred levels.

The truck business has also benefited from a large number of road test data.

Li Hengyu, the head of the truck division, made such a metaphor: “The virtual driver who is driving the car can be understood as taking C, the driver, the driver of the truck is taken, and there are some capabilities between the two. In fact, we took a C photo first, take A photo, which is very helpful for us. “

He believes that many underlying logic is the same in the automatic driving area, whether commercial vehicles or passengers.

For example, infrastructure, simulation system, operating system, perception, etc. are common.

Xiao Ma Lexing Co-founders and CTO Building Tiancheng have pointed out that Robotaxi and trucks are mutually promoted.

First, in terms of visual, all data on the way is identified by the sensor, which can be trained and learned.

Second, the truck is very different when driving the full goods and empty stocks, and the quality of the vehicle needs to be accurately controlled, which can be reflyed to the control accuracy of the truck.

3, Ponytron’s origin

Xiao Ma Zhi’s truck service is named “Xiao Ma Zhika”, English name “Ponytron”.

This English name has 3 mains:

First, with Chinese “creation” word tone;

Second, it is similar to the pronunciation of the truck English “TRUCK”;

Third point is more interesting, the group of the small horse truck team named “Seibertron”, and later TRON was naturally used as the element of the little Ma Zhika.

Xiaoma Zhun’s first announcement of the automatic driving truck information was at the Shanghai Auto Show in April 2019, when Xiaoma Zhi, announced the library to automatically drive long-distance freight fields.

However, in fact, the pony starts to develop trucks automatic driving technology in August 2018, which is more than half a year.

After the official official in the Shanghai Auto Show, in November 2020, Xiaoma Zhun first showed the L4-level automatic driving hardware and software system for heavy trucks.

Only one month, Xiao Ma Zhi Dynasties officially established.

Pony’s automatic driving truck team is led by Li Heiyu, vice president of Xiaoma Zhi, and the person in charge of the truck division.

Li is the first employee of the Xiaoma Beijing Office. After the completion of the Sichuan University Computer Department, the school enrolled Baidu and dominated the development of Baidu advertising budget control system, and won the annual bonus of millions of dollars in 2013. “Baidu’s highest award”.

After the establishment of the small Malaysian Truck Division in 2020, the first automatic driving truck test license for Guangdong Province has been obtained in December.

At present, Xiao Ma Zhikka is based on the back-mounted transformation of FAW liberation, and the road test in China and the United States, and the domestic test site is mainly located in Guangzhou.

Li Hengyu did not disclose the number of car to drive the truck, but he said:

“The automatic driving truck, if you are small, you need a certain size. At present, we are mainly developing research and development because of Tier 1, OEM, because the demand for automatic driving trucks is more clear.”

It is worth noting that FAW Liberation is also a strategic investment in Xiaoma Zhi, and FAW Liberation has also created a L3 level J7 automatic driving card in Joint Phase Technology.

4. In the next three years, the automatic driving system cost is reduced to the one-year income level of the driver.

Any business model, eventually returns to commercialization.

For Xiao Ma Yi, Robotaxi’s commercial path is relatively long, and the commercialization of automatic driving trucks is relatively faster.

The logistics industry does require automatic driving to solve some industrial pain points.

At present, my country’s logistics industry is huge, but there are problems with high accident rate, high cost, small scale and freight driver’s gap. In this context, many automatic driving companies aim at the automatic driving truck market.

Zhang Ning, a vice president of Xiao Ma Zhi, and Zhang Ning, head of the Beijing R & D Center, hoped that in the next three years, the cost of automatically driving to the driver’s annual income.

This is also equivalent, if a normal truck is transformed into an automatic driving truck, it is probably to recover the renovation costs within 1 year.

Xiaoma Zhiwang hopes to greatly improve the driving safety of the logistics industry, make up for labor insurance, and solve the problem of human management scale. The establishment of large teams will enhance the operational efficiency of the industry.

Pony publishes the timetable for automatic driving truck services in unmanned, massive and commercialization:

(1) Unmanned dimensions

Comparison road conditions technology research and development and testing in 2019 to 2020

2021 to 2022 realized a closed environment without driving

2023 to 2024 achieved limited scenario universal driving

(2) Gas production dimensions

Realization of a variety of vehicle platform after 2019 to 2020

2021 to 2022 Realization of pre-installation ranging from 2023 to 2024 Realization SOP

2025 realization automatic driving truck scale application

(3) Commercial dimension

Exploration on logistics and freight commercialization in 2019 to 2020

2021 to 2022 established a large-scale automatic driving team

2023 to 2024 established a smart logistics network

From this schedule, you can find that Xiao Ma Zhiqi will focus on the prototype design of the front mass production.

In the commercial chain of automatic driving, the Pony will play what role?

Li Hengyu said that the traditional truck business model has an OEM, Tier 1. The logistics business is formed throughout the commercial closed loop. Tier 1 provides truck parts, OEM production trucks, selling to logistics providers to obtain sales revenue, logistics providers get transportation services Operating income.

Future truck business model ecology is:

In addition to OEM, Tier 1, and logistics providers, the technical service providers such as Xiaoma Zhiwen cooperate with OEM and Tier 1 to provide automatic driving truck technology to all parties.

At the same time, because the automatic driving truck becomes a smart terminal, massive data can be generated, and smart logistics services can be produced.

It is worth noting that Li Hengyu said that in the future, Xiaoma Zhikong can purchase vehicles from OEM, build a smart logistics team, and provide smart logistics services.

At the same time, the smart logistics capabilities of the pony can also enplacement to logistics merchants.

Zhang Ning said that there are still many challenges that need to be resolved in the process of truly in accordance with the automatic driving technology of the car, the automatic driving technology above the vehicle.

“In the market of trillion dollars, only 100 levels (the team produced) is not reflected in the final business form.”

Li Lin Tao, head of Xiaoma Zhi, Beijing Hardware R & D team, said:

“The size of hundreds of vehicles is not our ultimate goal. We hope that in the future, we will gradually implement premium production, and hundreds are just the starting point. We need thousands of, we must have thousands of, forever and pretending.”

In addition to the automatic driving truck prototype technology, the current small horse has an important development work, which is a mining of massive road test data.

“How to use the data above 5 million kilometers is still a problem, and the throughput is not enough. If there is no good data digestion and absorption ability, then massive data may not be played at a tens of thousands of kilometers.

If there is no ability to drive under the road of complex urban roads, blindly stacked, and the vehicle will be interfered with other owners.

This track needs to be gradually progressing, constantly accumulating, the faster the power will run. Li Lin Tao said.

Leave a Comment

Your email address will not be published. Required fields are marked *