[? Yio guide?] ? The first thing you need to do is to get your hands on a new car, and then you'll be able to get it back.
Author: Yang Yaru
Editor: Zhang Yan
After the outbreak of the new coronary artery disease, a mysterious "courier" was sent to the streets of Wuhan to deliver the medical supplies to Wuhan's ninth hospital. This is the first single intelligent delivery in Wuhan. And to complete the distribution of this is a L4 automatic driving technology level of intelligent robots.
In areas where the condition is more serious, unmanned distribution by driverless technology has become a good way to replace human labor. Autopilot technology, which entered the winter in 2019, has subsequently become the center of attention once again.
The other side of the driverless breakthrough in unmanned logistics is - in the field of automatic driving technology for passenger cars, commercialization has entered a bottleneck, and a number of companies are difficult to sustain in 2019, heading for closure. Some industry insiders have begun to consider: if the L3 and L4 level of autonomous driving technology if the difficulty of mass production is too great, are there other solution routes? If the smart car can not immediately enter people's lives - "smart car" + "smart road", can it be feasible?
With policy support and drive, the "vehicle-road collaboration" technology route for autonomous driving has attracted more attention and discussion.
Intelligent Vehicle Infrastructure Cooperative Systems (IVICS) is the latest development direction of Intelligent Transportation Systems (ITS). It is the latest development of intelligent transportation system (ITS). By carrying relevant software and hardware on the car and road, it can carry out dynamic real-time information interaction between the car and the road.
Fu Yuwu, honorary chairman of the Chinese Society of Automotive Engineering, once said, "There is no smart city, no smart transportation, where is the smart car?" This also reflects the importance of vehicle-road collaboration, which should be treated as a core element of the continued development of autonomous driving.
Some industry analysts believe that the next 3-5 years will usher in the outbreak of vehicle-road synergy.
Whether it is ICT companies represented by Huawei, or automotive suppliers represented by Dexerials, Junsun Electronics, and vehicle-road collaboration solution providers represented by Hidi Smart Driving, all are actively participating in this track. In addition, BAT (Baidu, Tencent, Ali) is the most concerned players.
Baidu has built an open-source platform with its circle of friends accumulated over the years, and hopes to provide technical framework support through the open platform; Ali, which has an advantage in data, is exploring roadside scenarios; Tencent, which has also built an open-source platform, hopes to become an "ecological connector", and is carrying out the construction of the road informatization platform.
The giants don't want to miss any opportunity to bet on the future of transportation.
01
Baidu: Open Source Seizes Attention Portal
"'Smart car + smart road' will be the optimal solution for realizing autonomous driving." Baidu Vice President and General Manager of Intelligent Driving Business Group Li Zhenyu said, and this is also an important direction of the current Apollo force. Baidu believes that vehicle road collaboration can help 54% of the problems encountered by single-vehicle intelligence in road testing, and reduce 62% of the single-vehicle intelligence takeover.
In 2016, Baidu has begun to layout V2X vehicle-road collaboration, and as the lead unit of the national 03 special project "5G support ICT convergence of key technologies for autonomous driving" (expected to be completed in 2018), to carry out related research. However, it was not until September 2018 that Baidu officially joined the melee and launched the first open source program for vehicle-road collaboration.
In recent years, "open source" is Baidu's high-frequency word. Whether it's its deep learning framework PaddlePaddle or Apollo driverless system, Baidu small program, the keyword of open source as the core meaning, which of course also includes vehicle-road collaboration. In this way, Baidu hopes to help gain more favor with developers and build its own ecosystem.
Industry analysts believe that in this process, Baidu as a platform for the rise of the information portal hope to continue to grasp the attention of the "original driver" to compete for the next era of the ticket. For Baidu, the important thing is not who will produce driverless cars. Rather, it's that Baidu will be able to get the attention of its business partners in the first place when they launch their driverless cars.
In recent years, based on the four-layer open technology framework of the Apollo open platform, the Apollo Vehicle-Road Synergy open source has added or upgraded modules related to vehicle-road synergy at the level of software, hardware, and cloud services, etc. In December 2019, at the annual Baidu AI Developers Conference, Li Zhenyu once again reiterated the importance of vehicle-road synergy to the development of autonomous driving technology and the industry.
Apollo Vehicle-Road Collaboration Open Platform/Yiou Automotive Analyst? Yang Yaru
In order to attract more enterprises to participate in it, Baidu has even opened up the vehicle-road collaboration platform: the overall packaged edge intelligence and cloud capabilities and other services, to provide the Apollo platform developers with the ability to provide the infrastructure base, through the easy-to-call interfaces to support the intelligent information control, intelligent transportation, automated driving, intelligent parking, intelligent freight transport, intelligent automotive connection and other different industry application scenarios.
Baidu Vehicle-Circuit Collaboration has made concentrated efforts in three major directions, including the development of roadside sensing capabilities that meet the needs of autonomous driving scenarios. In addition, Baidu cooperates with communication chip and equipment manufacturers to optimize the V2X communication transmission channel for the needs of autonomous driving applications, and continuously accelerates the use of V2X roadside sensing information in the integration of automated driving systems on the vehicle side.
Up to now, Apollo has a variety of regional scenarios such as Beijing, Xiongan and Silicon Valley, as well as a variety of models such as passenger cars, unmanned minibuses, unmanned logistics vehicles, etc.; it has laid out and researched and developed full-stack vehicle-road collaboration technologies in terms of road-side sensing sensor solutions, road-side sensing algorithms, vehicle-side sensing fusion algorithms, data compression and communication optimization, V2X terminal hardware and software, and V2X safety. Apollo also has a certain amount of data accumulation in a series of scenarios such as unmanned fleet and open road unmanned vehicle test mileage.
On the road of vehicle-road synergy, Baidu is weaving a bigger net, and the scope of the circle of friends is also expanding, which will also become Baidu's first-mover advantage.
02?
Ali: Creating Scenario Advantages
Ali is making more attempts to create a "smart road".
During the Hangzhou Yunqi Conference on September 20, 2018, Alibaba Group formally announced the upgrade of its automotive strategy: to extend from the car to the road, and to create a "smart highway" using vehicle-road collaboration technology. This strategy will be accomplished by AliOS together with Dharma Academy, Gaode, Alipay, Qianxun Location, Zebra Network and other ****, together exploring the road in the next twenty years.
To this end, Alibaba Dharma Academy and the Institute of Highway Science of the Ministry of Transportation and Communications (hereinafter referred to as the Institute of Highway) signed a strategic cooperation to set up a joint laboratory for vehicle-road collaboration.?
Ali's intelligent vehicle-road collaboration solution for intelligent highway/Yiou automotive analyst? Yang Yaru
In May 2019, Ali invested 3.6 billion yuan in Thousand Square Technology (the main business of which includes intelligent transportation and intelligent security), and the two sides plan to carry out a number of collaborations in the field of intelligent transportation and edge computing, including technology research and development cooperation, marketing cooperation and project landing cooperation, etc., and **** the same to promote the intelligent transportation application solutions and the "cloud + edge + end" full chain of solutions. edge + end" full link solution implementation. This again confirms the importance of Ali for intelligent transportation and the field of vehicle-circuit coordination.
Specifically, Alibaba's vehicle-road coordination strategy is mainly planned to promote from the cloud control platform, intelligent perception base station, cooperative computing system at three levels. In the road side side can use the Dharma Institute perception base station technology, AliOS is responsible for building the car perception and collaborative computing, cloud control platform plays the role of the cloud brain.
The Tmall Elf in-car voice assistant in the cockpit/AliOS website
In Ali's planning, the core component of the road-side of vehicle-road collaboration is the perception base station. That is, composed of a variety of sensors and high-efficiency computing units, can help the vehicle route planning and positioning of the brain outside the car. At present, Ali has with Intel, Datang Telecom Group to carry out intelligent transportation - vehicle-road cooperation in the field of strategic cooperation. Based on the LTE-V2X communication technology and edge computing capabilities of both parties, AliOS hopes to promote the commercialization and standardization of the entire system by building an ecosystem and introducing more application scenarios.
Because of the identity of Ali's e-commerce platform, Ali's main advantage is currently focused on the road side, and its development of vehicle-road coordination has the accumulation of data such as the Rookie Bird Alliance scene and the ET City Brain.
In the closed road-side scenario, Ali also has a comparative advantage. Industry sources analyze, from the current situation, Ali may be in the freight logistics industry can be the first to land on the automatic driving, if for vehicle-road cooperation for in-depth research, fixed logistics routes are relatively easy to transform, the use rate is also relatively high. With this, Ali may be able to greatly reduce marginal costs in the future.
03?
Tencent: play "ecological connector"
In recent years, Tencent has been betting on the field of intelligent travel, has also begun to move.
In May 2019, Tencent announced that it would unite with Intel, Nokia, China Unicom, and Neusoft to establish a 5G vehicle and road collaboration open source platform, which has already been open sourced for vehicle and road collaboration and micro-service frameworks.
Vehicle-road collaboration innovation ecology/Yiou automotive analyst? Yang Yaru?
Inseparable from the significance of the attributes as a connector, Tencent this platform is more like a connector between the application and the network, by releasing the potential of the 5G network, connecting people, cars, roads and clouds. Tencent wireless operation department Zhang Yunfei introduced, "At present, the car enterprise and the road side invested a lot of resources for vehicle road collaboration transformation, but ultimately did not open up the C end, failed to realize the overall closed loop of the information flow and business flow, as well as commercial realization. Now the information of the road shows a trend of fragmentation, through such a platform can realize the relevant connection, help the development of road informatization and the road towards intelligence."
In order to facilitate the use of cooperative users, the Tencent platform uses micro-services through the integration of physical machines, virtual machines, containers and other ways to realize the management of third-party application services and massive node management issues, and supports a variety of programming languages. In addition, Tencent also provides three integration capabilities, including user interface integration methods, interface integration methods, and development framework integration methods.
Previously, Tencent future network lab expert Yu Yifan revealed that Tencent 5G vehicle-road collaboration open source platform for edge computing, vehicle-road collaboration business itself provides a series of modules, covering the road-side edge computing platform, the software system and the 5G network docking flow control, to CV2X message set as the core of the data processing function, etc.. In addition, the platform can upwardly support services in the transportation industry related to road management, traffic control, etc., and downwardly can be docked to V2X and other infrastructure construction.
Up to now, Tencent has built a series of core capability systems for the vehicle-road coordination platform through self-research. In the field of visual perception, Tencent has a specialized research and development team is researching a number of key capabilities, including image recognition, multi-sensor fusion, fusion decision-making, and modification of vehicles in complex environments, such as testing; between the car and the car, the car and the road communication, Tencent force 5G standards, the establishment of the future network laboratory, mainly to study the **** problem that will be faced by the landing of the 5G vehicle-road cooperation; in terms of landing, Tencent is Consider strengthening the deployment of edge computing devices, is currently working with a number of partners to jointly build edge computing open source platform.
Based on its own characteristics, Tencent has created a unique path that is more suitable for itself, and has well connected the entire industry.
04
Who is the "temporary leader"?
Baidu, which has been working on Apollo for a long time, seems to be the "temporary frontrunner" at present. Baidu seems to have a clear first-mover advantage, with an established ecosystem and extensive open-source collaboration.
Baidu vehicle-road coordination solutions show / Baidu official website video screenshot
Up to now, Baidu has dozens of car companies have reached a cooperation, and with a number of provinces and municipalities in the region of the city to start the relevant cooperation. In the vehicle-road collaboration scenario, Baidu has already made a mark in unmanned vehicle testing methods, scenario construction, vehicle networking, high-precision maps, etc., and its V2X non-technology has also been laid out earlier. But at the road end, Baidu still has a short board.
And on the roadside, operational scenarios, etc., is Ali's advantage.
Unlike Baidu's open road, Ali is launching more exploration based on its own business accumulation, and with the support of the logistics advantage of the Caijiao alliance, Ali is likely to make breakthroughs for closed test scenarios. Mature scene model, is conducive to accelerate the rapid landing of vehicle-road coordination.
At the end of 2016, Ali began the exploration of "Internet + transportation". Now, the Dharma Institute has also become an important hand in the layout of Ali's perceptual base station. On the other hand, although the road side has been explored, but the coordination of roads and vehicles to bring multi-sensor fusion and real-time transmission of signals is more difficult, which is also the future of Ali to face.
Some industry insiders have told Yio that Ali still seems to be hesitant about how to promote vehicle-road coordination. In addition, constrained by the profitability problems faced by the car road synergy, Ali's attitude is also more cautious.
In comparison, Tencent's advantage is to understand the user. It is also building a vehicle-road collaboration ecosystem through the construction of three aspects, including infrastructure, platform services, and business applications. At the same time, there are also a lot of Tencent Department of intelligent network technology enterprises involved in the cooperation of the local car road cooperation. But for Tencent, the future still needs to improve the core competitiveness and let the outside world perceive the advantages of Tencent's related technology.
The complexity of the scenarios, the length of the industry chain, and the vast scope of the industry are all roadblocks in front of the development of the industry. Any one company can not rely on their own strength to deal with a series of software, hardware, platform, construction, application and other work of vehicle-road coordination. Some industry insiders told Euromobile that with the development of the industry, the opportunities of the future segmentation of the industry still exist, and some startups are also expected to get a share of the pie.
05?
Waiting for the wind to come
Vehicle-circuit coordination may enter a period of rapid development.
From the perspective of the general environment, China has a unique opportunity in the exploration of vehicle-road collaboration. According to incomplete statistics, there are more than 30 demonstration zones in the country, including 10 national demonstration zones in Shanghai, Beijing-Hebei, Chongqing, Wuxi (pilot zone), Hangzhou-Tongxiang, Wuhan, Changchun, Guangzhou, Changsha, Chengdu and so on. The test scenarios are also gradually moving from closed to open, from single to multiple, and from urban to high-speed.
The moving car/Unplash
Yiou Wisdom once predicted that 2019-2021 will enter the introduction period of vehicle-road collaboration industry development, especially road upgrading will become the focus of urban layout, and the layout represented by demonstration zones will gradually form a scale.
For the host factory, vehicle-road synergy reduces the difficulties in manufacturing to a large extent.
Volkswagen (China) Investment Co., Ltd. Cai Jiyong once said, as long as the car end with L2 level of single-vehicle intelligence, you can realize the high-level automated driving vehicle road coordination. At present, many car companies have been betting on L3 level, L4 level of automatic driving, greatly increasing the cost of research and development, and the car's hardware investment.
Many industry players seem to have formed a **** knowledge: vehicle-road synergy will usher in a period of concentrated outbreak, perception level or the first to usher in the development. But the difficulties still need to be resolved.
First of all, the infrastructure level, due to vehicle-road coordination relies on communication technology, through remote control to realize high-level automatic driving must rely on 5G standard landing. At present, whether the road system can follow the 5G standard landing, synchronized realization of hardware for automatic driving support, is not yet known.
Secondly, from the point of view of the current situation of vehicle-road network, the development of various aspects is not balanced. Some experts have proposed that vehicle-road collaboration should be systematically constructed in order to pull the three-party construction into the right track.
Thirdly, standards are still missing - which will also become an important factor limiting development.
Vehicles on the road/Unplash
And for enterprises, the business model of vehicle-road collaboration has become an important discussion topic.
Multiple practitioners have expressed concerns to YEEO Automotive that the business model of Vehicle-Road Collaboration is not obvious at the moment. Zhao Shengqiang, senior solution architect of AliCloud Universal Competence Center, said that at the moment, although various enterprises have launched relevant research and exploration of vehicle-road collaboration, at present, the entire industry has not found a clear business model, but also need time for trial and error.
Edited by Yan Zhang
This article comes from the authors of Automotive Home, and does not represent the viewpoint of Automotive Home.