Introduced Tokyo Unmanned Vehicle Technology This company staged an automatic driving curve overtaking


In 2018, Google’s Waymo had to be commercialized. The start-up company that was established at this time had a play in the auto driving business.

The answer for this Shenzhen company is that there is a play, as long as it goes directly to the mass production road.

On May 11, China's Shenzhen Automotive Electronics Co., Ltd. Golden Language Technology and Japan's veteran autopilot research and development company ZMP, announced a joint venture to establish the autonomous driving company "Shi Air Intelligence," and will be launched this year, production models of L4 unmanned logistics Trolley.

Dare to declare that it has entered the mass production phase within half a year of establishment, and it is still the first time in the field of domestic automatic driving. The car and the intelligent founding team of Shih Air communicated in depth and found out that this company dared to enter the Red Sea when it started driving itself, and the secret to the rapid mass production of its products.

I. Escaping Intelligence: Using Japan's 10-Year Autopilot Accumulation in China

Mr. Ren Yi, CEO of Shih Air Intelligence, established Jinyu Technology in Shenzhen in 2009 to develop automotive electronic products. The main products are car navigation and car machines.

东京奥运会无人车,诗航智能

â–² poetry aviation intelligence CEO Ren Yi

In the course of nine years of development, Golden Language Technology has entered the once-closed Japanese automobile supply chain, becoming one of the mainstream brands in Japan's automotive navigation electronics industry, and also one of the few Chinese Tier-2s entering the Japanese automotive supply chain.

东京奥运会无人车,诗航智能

â–² gold language technology product line

However, Ren Yi, the automotive industry person who used Jobs as his entrepreneurial icon, has always been looking forward to the technologies that can change the lives of people. He is also eager to tackle the topic of AI and auto driving.

By chance, Ren Yi met with President ZMP of Japan’s autonomous driving company, Gu Guheng, last year. ZMP is the company that wants to build a driverless taxi for the Tokyo Olympics.

东京奥运会无人车,诗航智能

â–² Taniguchi

Founded in 2001, ZMP started with the development of robots. In 2008, ZMP began autopilot-related R&D work. In 2014, ZMP's autopilot test vehicle got on the road and received investment from Intel. Last year, the company obtained permission from the Japanese government to begin testing autopilot on the open road.

In the ten years of R&D accumulation, ZMP has formed its own autopilot technology and products - a binocular depth camera that provides automatic driving awareness, and an on-board computer "IZAC" that can realize the perception, planning, and control of autopilot.

东京奥运会无人车,诗航智能

â–²The picture above shows the ZMP depth camera RoboVision. The picture below shows the ZMP automatic driving brain IZAC.

东京奥运会无人车,诗航智能

The ZMP modified self-driving taxi has been able to run smoothly on the open road in Japan.

东京奥运会无人车,诗航智能

In addition, ZMP also launched CarriRo Delivery, a self-driving logistics vehicle.

东京奥运会无人车,诗航智能

In the field of automatic driving, ZMP already has a relatively complete set of technologies and product systems.

However, ZMP is a company that focuses on R&D and light-marketization. Japan's domestic deployment of autonomous driving technology is also relatively conservative. Therefore, despite the long-standing accumulation, ZMP has not been able to win too much in the commercialization of autonomous driving.

Ren Yi saw the opportunity - the introduction of ZMP's technology into the country, and the application of localization. ZMP is also happy to be able to open up a huge market like China. Both parties hit it off and decided to start cooperation.

Jinhua Technology and ZMP's joint venture, Poetry Air, was founded. Japan provides basic technology, and China is responsible for localization and commercialization of technology.

Second, the logic of passing the curve: direct killing to landing

However, the situation of self-driving entrepreneurship is not favorable for Shih Air Intelligence at this time. Hundreds of companies with surnames and surnames are influential, and this track has long been used to establish the Red Sea. From ADAS to L4 and above, there are company coverage.

Where are the opportunities for the poetic aviation intelligence that was established in 2018?

The best way to postpone strikes is to use force and use ZMP to accumulate ten years of technology and quickly enter mass production.

On the road map for development, Shih Air Intelligence set a rhythm like this:

In 2018, the L4 self-driving car was launched for testing in the country. Automatic logistics vehicles and automatic sweepers were introduced. In the same year, binocular cameras and ADAS products were installed.

In 2019, the L4 automatic driving long-distance logistics vehicle was launched. At the same time, a pre-installed ADAS solution for the host plant was introduced.

东京奥运会无人车,诗航智能

Song Zhen intelligent CTO Liu Zhenyu, who has 10 years of experience in the automotive electronics industry, tells the car that they are taking the route of an L4 commercial vehicle's autonomous driving and the ADAS program. The most central strategy for choosing these two routes is to be able to land first and make money first.

东京奥运会无人车,诗航智能

â–² poetry intelligence intelligent CTO Liu Zhenyu

In the direction of the ADAS, Liu Zhenyu said that although ZMP’s previous R&D focus was not driver assistance but driverless driving, the underlying algorithms, such as object recognition and lane line recognition, were similar. This means that the bottom R&D work of the poet Airways in the direction of the ADAS can be greatly reduced and it can be quickly commercialized.

On the L4 route, Liu Zhenyu chose to focus on the logic of commercial vehicles to the analysis of vehicles:

In terms of security, the existing autopilot technology is still difficult to guarantee the required safety level of passenger cars, while commercial vehicles are less concerned with the safety of passenger vehicles, and thus will be more easily accepted by the government and accepted by the industry.

Technically, ZMP has a long history of road test data and technology accumulation is an important advantage. On the other hand, the automated logistics vehicles and sweepers that Shihhung Air Intelligent will exert are under low-speed working conditions for most of the time, and the technical difficulty is even less. The automated long-distance logistics vehicles, which run for most of the time, are structured. On the road, working conditions are simpler and technically easier to implement.

In terms of cost, the commercial vehicle sector is currently more able to accept the cost increase from the application of autopilot technology than passenger vehicles.

The superposition of various factors makes the L4 commercial vehicle drive automatically, making it the best choice for high-grade autopilot in the pursuit of mass production speed.

After selecting an easy-to-implement autopilot scenario, in terms of implementation methodology, Shih Air Intelligence also chose to serve mass production speeds – using mature solutions that look as low as possible.

On the automated driving computing platform, although many autopilot start-up companies are using Nvidia's Drive PX2, Liu Zhenyu said that the poetic intelligent autopilot program will not use it, but will rely on ZMP's autopilot brain IZAC. In addition to cost and power consumption, Liu Zhenyu believes that the biggest problem lies in the product life cycle. Drive PX2 does not guarantee its life cycle, and the electronic components in the car often require several years or even ten years to work normally.

The ZMP IZAC is a car-level product. This means that there is no need for long-term verification work on the computing platform, which is the most important hardware for autonomous driving. "I took it and put it on the train." This is Liu Zhenyu's "guarantee."

However, the car thing concerned that the IZAC's main computing chip is an Intel's low-end processor Celeron N2930, in terms of computing power does not seem to meet the L4 automatic driving demand. Liu Zhenyu explained that the sensor data fusion that dominated computing power was placed on an FPGA, and IZAC was running a core algorithm, so the current power is not a problem.

According to the car thing, the logistics car that Shihei Air Intelligence launched on the basis of CarriRo Delivery this year will be operating on a fixed route. At present, it has already obtained intentional orders from two large domestic Internet companies, and both companies are developing their own autopilots. Logistics trolley.

The L4 self-driving long-distance logistics vehicle to be launched next year by Shihang Aviation Intelligence will adopt the automatic driving mode on the high-speed road, and it will be taken over by the human driver after the next high-speed road.

In order to meet the development needs of the two technical routes, Shih Air Intelligence has drawn some talents from Golden Language Technology, and ZMP has also stationed technical personnel in Shenzhen. This year, Poetry Air Intelligence will expand its team to 50 people.

Third, based on the local positioning of automated driving solutions

By using Japanese technology for its own use, Shih Air Intelligence can quickly enter the stage of commercialization.

However, in the area of ​​autonomous driving that is related to national security, will the identity of Sino-Japanese joint ventures adversely affect the development of Shih Air Intelligence in China?

For issues of major origin, Ren Yi and his team responded in two ways.

First of all, in the equity relationship of the joint venture, Shih Air Intelligence is controlled by the Chinese side, and the proportion of the equity of China and Japan is 60%: 40%, which guarantees China's control over the management of the company.

Secondly, in the negotiations with ZMP, Shihang Intelligent won as much as possible the ZMP underlying patent authorization to ensure the mastery of its technology.

“The poetic airline is not ZMP China. We will introduce Japan’s technology and make it a secondary development. So we’re not shy away from cooperating with Japan and don’t hide it.”

At present, there are two important tasks for Shih Air Intelligence to localize the ZMP technology. First of all, the technology is applied to different models, and the requirements for the sensors will change. The team of Shih Air Intelligence needs to recalibrate and adjust the sensors. Secondly, China’s road conditions and traffic rules are also quite different from those in Japan. For example, China is a left-handed country, and Japan is a right-hand rudder. At the same time, China’s road conditions are more complex. This requires the collection of a large amount of local data collected by Poem Airways’ intelligent algorithm. Make changes and enhancements.

In addition, automatic driving with L4 or higher requires high-precision maps. In this regard, Shih-Air Intelligent will cooperate with domestic map dealers to solve the problem.

In order to realize commercialization in selected areas as soon as possible, Poem Air Intelligence will not expand its area of ​​involvement too much. In the business model, although many start-up companies are attempting to become operators of unmanned fleets, trial flight intelligence does not intend to go this route for no reason. The model is too heavy, it consumes too much money, and the cycle is too long.

The positioning of Shih-Hung Intelligence on itself is that the solution provider of autonomous driving technology provides auto-driving hardware, software, and supporting services, and the operation work will be handed over to downstream companies in the industrial chain.

In addition, Liu Zhenyu also disclosed to the car that, in the ADAS program for the front-loading market for passenger cars, the poem Airways Intelligence has reached cooperation with a domestic Tier-1, and obtained the latter’s investment.

Ren Yi and Liu Zhenyu, born in the automotive industry, believe that as long as they face the front loading market, there is no possibility of bypassing Tier-1. Therefore, in the ADAS field, the poet Airways positions Tier-2.

Concluding remarks: The soil in which poetry voyages were established - automatic driving calls to land

In 2018, more and more self-driving entrepreneurs expressed their perceptions to cars: investors prefer to ask when they can make money. Even Google's Waymo, all of them are constantly making announcements to commercialize this year. In other words, the industry is currently focusing on technology.

The introduction of technology from Japan and the easiest path to commercialization have all reflected the characteristics of this Shenzhen company—pragmatic, or how to turn technology into money as quickly as possible.

It should be noted that the establishment of a company such as Shih-Hang Intelligence reflects the arrival of such a node. The industry's expectation for the application of autonomous driving technology has far exceeded the focus on technological development.



Aluminium Machining Parts

Aluminum machining parts are those aluminum parts that made by cnc machining processings. They could be alumnum pure machining cooling plates or machining based on aluminum cold forming, extrusionsor forgings.

Aluminum parts are widely used for various industries, and they are with features of easily processing, flexible to make different complicated shapes and suitable for different surface treatment.

We have many years manufacturing experiences on aluminum alloy heatsink parts
our mature technology on extrusion and machining make our Heatsinks components brand very popular at our domestic market. And we also export to EU, US and other Asian markets. With strong engineering team on extrusion profile radiators and metal machining components, we have been provided our customers very excellent services many years and established very stable business cooperation relations.
aluminum Machining platePreision aluminum Machining PlatesMachining Aluminum RingsAluminum cooling Plate

Aluminium Machining Parts,Aluminium Cnc Turning,Aluminum cooling plate, Aluminum cooling plates

Suprel Metalwork Co., Ltd , https://www.suprel-cool.com