Huawei will lay out how the smart car industry will develop in the future

Huawei is developing intelligent connected cars. The Internet is influencing people's way of thinking and interaction at an unprecedented depth and speed. As an important field that is indispensable in daily life, the automotive industry is also undergoing major changes in product form, technology path and development direction with the changes in the Internet era.

With the continuous advancement of autonomous driving and vehicle networking technologies, the prospects for smart cars are becoming increasingly clear. Many leading companies in the industry have crossed over to the automotive field and want to get a piece of the automotive market. As the world's leading ICT (information and communication) infrastructure and smart terminal provider, Huawei has inherent advantages in the technology and Internet networks required by smart connected cars, and is therefore gradually entering the automotive field.

1.1 What happened to Huawei this time?

The intelligent vehicle Sharing-VAN created by Huawei and Dongfeng Motor will be unveiled at the Shanghai Auto Show in late April. Huawei's layout in intelligent connected vehicles has begun to bear fruit.

Sharing-VAN integrates the latest technological achievements such as LTE-V/5G, driverless driving, new energy electric vehicles, time-sharing leasing, and public transportation services. It is an important step taken by Huawei in the field of intelligent connected vehicles.

Huawei mainly provides technical services such as in-vehicle communications, 5G and cloud data centers to help realize functions such as information query, traffic dispatch, vehicle management and equipment management.

As early as October 2014, Dongfeng and Huawei Auto signed a strategic cooperation agreement, planning to build emotional self-driving smart cars in three steps: car machine, car networking, and smart cars. The Sharing-VAN, the first driverless car integrated with LTE-V/5G, is an important product in the strategy. The cooperation opportunity between Huawei and Dongfeng stems from Huawei's optimism about the future of smart connected cars and its confidence in its own technology matching.

1.2 Huawei's layout in the automotive field

Huawei's layout of intelligent connected cars in all aspects of chip-device-pipe-cloud. Huawei's layout of the automotive business can be traced back to the establishment of the "Internet of Vehicles Business Department" in 2013. Subsequently, relying on LTE-V and 5G technologies, it has evolved from focusing on the Internet of Vehicles and intelligent driving business to the entire Internet of Vehicles ecological industry chain.

Huawei owns the world's first multi-mode chip Balong 5000 that supports V2X, and also has the most advanced 5G technology. It is also one of the standard initiators of C-V2X. In the cloud, it has a combination of OceanConnect IoT platform + Huawei public cloud. Therefore, Huawei has become the only C-V2X vehicle networking manufacturer in the world that provides a full range of solutions from chip-end-pipe-cloud.

2. Huawei’s entry into the market shows the broad prospects of intelligent connected vehicles

The prospects for intelligent connected cars are broad, as can be seen from Huawei's low-key entry. Intelligent connected cars are an organic combination of vehicle networking and smart cars. By equipping on-board sensors, controllers and other devices, and integrating modern communication network technology, information exchange and sharing between cars, cars and people, cars and roads, and cars and backends can be realized, achieving safe, comfortable, energy-saving and efficient driving. According to BIIntelligence's forecast, by 2020, global sales of connected cars will increase from 21 million in 2016 to 94 million, accounting for 82% of the market, and the market size of intelligent connected cars can reach 100 billion yuan.

The continuous expansion of downstream scenarios and policy support are driving the realization of this target scale.

In the future, we will enter a new intelligent era, where all things will be sensed and connected, which will be the basic characteristics of the future society. According to Huawei's GIV (Global Industry Vision), with the continuous update of downstream demand scenarios, the number of global connections will reach 100 billion by 2025, and the automotive industry will also benefit from this wave of change and transform to intelligent connected cars.

In December 2018, the Ministry of Industry and Information Technology issued the "Vehicle Network (Intelligent Connected Vehicle) Industry Development Action Plan", which clearly expressed the need to strengthen policy support for intelligent connected vehicles. The plan provides guidance that the development of the vehicle network industry will move from the demonstration application stage to the large-scale application stage by 2020. Intelligent connected vehicles with high-level autonomous driving functions and 5G-V2X will gradually achieve large-scale commercial applications, and "people-vehicle-road-cloud" will achieve a high degree of coordination, and the people's growing demand for a better life will be better met.

The growth of intelligent connected vehicles benefits from the development of both the Internet of Vehicles and intelligent driving, and its market will expand with the expansion of the Internet of Vehicles and intelligent driving markets.

2.1 Internet of Vehicles

The Internet of Vehicles is an important support for intelligent connected vehicles. According to the definition of Huawei's C-V2X white paper, the Internet of Vehicles is to use the new generation of information and communication technologies to achieve all-round network connections within the car, between cars, between cars and roads, between cars and people, and between cars and service platforms, improve the level of automobile intelligence and autonomous driving capabilities, and build a new form of automobile and transportation services, thereby improving traffic efficiency, improving the driving experience of cars, and providing users with intelligent, comfortable, safe, energy-saving, and efficient comprehensive services.

2.1.1 Development History of Internet of Vehicles

The first stage was from 1996 to 2015, which was the in-vehicle information service stage. In 1997, the prototype of the Internet of Vehicles, Telematics (in-vehicle information service system), was born in the United States. In 1995, GM launched OnStar as an exclusive configuration for its own cars to provide in-vehicle information services. At the China International Internet of Things Conference held in Wuxi in 2010, the concept of the Internet of Vehicles was proposed.

The second stage is from 2015 to 2025, which is the stage of gradually expanding from the Internet of Things to the Internet of Vehicles, and gradually realizing intelligent connected cars. At present, the Internet of Vehicles only realizes the interconnection between cars and the cloud, but the overall car networking rate is very low. However, according to Gartner's forecast, the number of connected cars in the world will exceed 250 million by the end of 2020.

The third stage is after 2025, which belongs to the smart travel stage, and the goal is to achieve car-to-everything connection. As the penetration rate of the Internet of Vehicles increases, cars and cars, cars and people, cars and roads, and cars and X will eventually be connected, and finally achieve full coordination of cars, people, and roads. At that time, drivers will be completely liberated, and the safety of autonomous driving will be fully guaranteed.

2.1.2 The three drivers of terminals, networks and clouds create a vast market for Internet of Vehicles

The expansion of the Internet of Vehicles market can be driven from three directions: terminal, network, and cloud.

The terminal refers to the intelligent sensor component of the car, which is used to collect vehicle operation information, driving status and surrounding environment. At present, it is constantly expanding from basic sensors to integrating artificial intelligence, voice recognition, vehicle-mounted systems and other fields to develop into intelligent cars. With the growing demand for car safety, comfort, communication and entertainment, the quantity and quality of various sensors and vehicle-mounted terminals have also increased. The number of sensors and vehicle-mounted terminals involved in each car has reached more than 200, and this number is still growing at an average annual growth rate of 7.3%. The market derived from sensor components is huge.

The pipe refers to the core infrastructure of the Internet of Vehicles - the network. In order to achieve external communication of vehicles, the two major technology camps of V2X (Vehicle to Everything) are DSRC and LTE-V. In the future, vehicles will realize autonomous driving, and the requirements for the network will become higher. The data generated by smart cars in one second can reach 1GB, which requires a matching network to synchronize the data to the cloud. The peak theoretical transmission speed of 5G can reach 10GB per second, which is 100 times faster than the transmission speed of 4G network. Therefore, only 5G can meet the network requirements of autonomous driving. LTE-V in the 5G era will evolve into NR-V2X. According to ABI Research, by 2025, there will be 50.3 million 5G-connected cars. The typical replacement cycle of cars is 7 to 10 years, so connected cars will increase significantly between 2025 and 2030.

The cloud refers to the collection of massive data from the Internet of Vehicles and the realization of data calculation, storage, processing and sharing. The Internet of Vehicles ecosystem includes content, software, equipment providers, telecom operators, automobile manufacturers and users. The cloud will collect massive data and require cloud computing functions such as virtualization and massive storage to attract more cross-border service providers to join and expand the Internet of Vehicles.

The growth of terminals, networks and clouds has created a vast market for the Internet of Vehicles. According to foreign media statistics, the market value of the global Internet of Vehicles market was approximately US$66.075 billion in 2017, and it is expected to increase to US$208.107 billion by 2024, with a compound annual growth rate of approximately 18.00% from 2018 to 2024. The domestic Internet of Vehicles market is also developing rapidly. Internet of Vehicles-related projects have been included in national key projects, with initial investment exceeding 10 billion yuan. According to the data statistics of the Prospective Research Institute, the space for the Internet of Vehicles market can reach a trillion scale within 5-10 years. According to the forecast that China's car ownership will be approximately 250-300 million in 2020, assuming that the penetration rate of connected cars is 15%, the number of vehicles with networking capabilities will reach about 40 million. Estimated at a hardware product price of 1,000 yuan per vehicle, the hardware market alone has a scale of 40 billion yuan.

2.2 Autonomous Driving

Autonomous driving is the highest stage of smart cars, and the continuous development of autonomous driving technology has driven the growth of smart connected cars. Autonomous driving cars rely on artificial intelligence, visual computing, radar, monitoring devices and global positioning systems to work together to allow computers to automatically and safely operate motor vehicles without any human active operation. Autonomous driving is an important development direction of smart connected cars and the highest stage of smart cars. To achieve autonomous driving, it must be carried out under the intelligent connected system.

According to SAE standards, autonomous driving is divided into six levels according to the degree of intelligence and automation: no automation (L0), driving support (L1), partial automation (L2), conditional automation (L3), high automation (L4) and full automation (L5). Currently, most companies are in the early stages of L2/L3. It is expected that in 2020/2021, leading manufacturers will take the lead in entering the more advanced stage of L4/L5.

2.2.1 Rapid penetration of ADAS drives autonomous driving forward

ADAS is the prerequisite for realizing autonomous driving. Before autonomous driving is fully realized, many unmanned driving technologies have been commercially applied as assisted driving technologies, and ADAS is one of them. ADAS (Advanced Driver Assistance System) is an active safety function integrated control system that collects data about the surrounding environment of the car through sensors such as radar, identifies and tracks objects, and makes decisions based on map data so that the driver can be aware of possible dangers and can directly control the vehicle when necessary. In the various stages of autonomous driving, L0-L5 belongs to the popularization and promotion stage of ADAS, and the realization of L3-L5 requires the combination of vehicle networking technology. At present, L1 and L2 have been commercialized, and KPMG predicts that L3, L4/L5 will be commercialized in 2019 and 2025 respectively.

The growth rate of the ADAS market is impressive. According to Goldman Sachs research statistics, the global ADAS penetration rate is generally not high. In 2015, the penetration rate in Europe, the United States and Japan was only 8%-12%, and it was even lower in China. However, through the continuous refinement of terminal scene requirements, the active research and development of car companies and cross-border Internet companies, the ADAS market will usher in rapid growth in recent years. Goldman Sachs predicts that the global ADAS market size will grow at a compound annual growth rate of 42% from 2015 to 2025. According to IHS estimates, the domestic ADAS market will exceed US$70 billion in 2020, and the CAGR will also exceed the global growth rate. The market and global share will expand rapidly.

As the penetration rate of ADAS increases, the market for autonomous driving also expands. ADAS, as a technical touchstone for autonomous driving, represents the early stages of autonomous driving. Its increasing penetration rate not only reflects the improvement of the autonomous driving market, but also implies the update of the autonomous driving stage. Combined with the demand for automation, safety, and environmental protection in future scenarios, the autonomous driving car market will continue to grow. According to Allied Market Research's research and statistical estimates, the global autonomous driving car market value will reach US$54.23 billion in 2019 and will increase to US$556.67 billion in 2026, with a compound growth rate of 39.47%. [page]

3. Intelligent and connected vehicles bring new opportunities for electronic devices

In the era of intelligent and networked cars, automotive electronic devices are welcoming new growth opportunities, and the market size of automotive semiconductors is expected to grow steadily in the long term. According to the latest "Intelligent Car Innovation and Development Strategy" (draft for comments) of the National Development and Reform Commission, by 2020, the technical innovation, industrial ecology, road network facilities, regulatory standards, product supervision and information security system framework of China's standard intelligent cars will be basically formed, the proportion of new intelligent cars will reach 50%, and medium and high-level intelligent cars will be commercialized. According to IHS forecasts, China's passenger cars will reach 27.7 million in 2020, so China's intelligent cars will exceed 13 million, and automotive electronic parts will continue to benefit from the rapid growth of intelligent driving.

In the era of intelligent driving, "on-board computers" and "on-board servers" are the general trend. Establishing a large closed loop of "sensing-fusion-decision-execution" will drive new growth in automotive electronic devices. In intelligent driving, if it is not possible to make intelligent decisions in time and control the direction to avoid obstacles when an obstacle is detected, but the command is transmitted to the cloud first and then sent to the vehicle terminal, a slight delay due to signal transmission and other reasons will lead to an accident. Therefore, a local assisted driving/autonomous driving control system with high-performance computing capabilities is required to fuse and process the data received by the sensors. "On-board computers" and "on-board servers" will be the general trend, forming a large closed loop of "sensing-fusion-decision-execution". Based on the above framework, we further analyze automotive sensors, microcontrollers, memory and PCB.

3.1 Sensor

The penetration rate of autonomous driving is gradually increasing, the base of the automobile market is huge, and the sensor market is expanding. Automotive sensors are equivalent to the eyes of autonomous vehicles, which can identify people, vehicles, obstacles and basic transportation facilities. The penetration rate of autonomous driving is gradually increasing, the base of the automobile market is huge, and the sensor market is expanding. Automotive sensors are mainly divided into three categories: lenses, ordinary radars, and laser radars (LIDAR).

LiDAR: It can detect the distance of the surrounding environment with laser pulses and draw 3D maps with software, thus providing enough environmental information for self-driving cars. The ranging accuracy of LiDAR is very high, basically reaching plus or minus one or two centimeters, or even millimeter level, and the resolution is also very high, but the cost is relatively high. Self-driving cars are an important driving factor for the growth of the LiDAR market. According to YOLE development, the annual revenue of the global LiDAR market will increase from US$726 million in 2017 to US$5 billion in 2023, with a CAGR of 43%. According to the statistics of the Forward-looking Industry Research Institute, the scale of China's LiDAR market reached 325 million yuan in 2017 and will reach 464 million yuan in 2022.

Car camera: Due to the high cost of LiDAR, traditional radar and car camera are its main substitutes. Car camera is the basis for realizing many warning and recognition ADAS functions. The processing of visual images is more intuitive for drivers, so car camera is indispensable. According to a research report by Tokyo Shoko Research, the global market demand for car lenses was about 23 million units in 2016, and it is expected to reach 80 million units by 2020. According to a report released by IHS Automotive, a professional automotive research company in the United States, the compound annual growth rate of car cameras in China from 2015 to 2020 will exceed 30%. With the advent of the era of fully automatic driving after 2020, the car camera market will show geometric growth.

Single vehicle sensor breakdown: Let's look at the sensors on a single vehicle. For example, the Tesla Model 3 uses a radar and eight cameras, which can only achieve Level 2 autonomous/assisted driving. It is conservatively estimated that a single vehicle needs to be equipped with at least 30 image sensors to achieve Level 5 autonomous driving. In addition to image sensors, a large number of pressure, temperature and other sensors will also be used in the powertrain system. It is estimated that in 2021, the shipment volume of sensors in the powertrain system will reach 1.8 billion. Calculated at $1 per sensor, the corresponding market space is conservatively estimated to be close to $1.8 billion.

3.2 Microcontroller

The prospect of automotive MCU is broad. As the automotive market turns to intelligence and networking, new requirements are put forward for the performance, safety, scalability, update and upgrade, connection, and low power consumption of MCU. We estimate that the number of microcontrollers used in a whole vehicle is about 36 to 54 based on the statistics of single vehicle splitting, taking into account the needs of safety applications, body control, power system, and battery pack. Considering that the unit price of automotive-grade chips is generally high, the value of the whole vehicle MCU is about 100 to 500 US dollars, calculated at 3 to 10 US dollars per chip. Based on the 27.7 million passenger cars in China in 2020 and the intelligent driving penetration rate of 50%, the market for intelligent driving automotive microcontrollers in China alone will reach 1.38 billion to 6.925 billion US dollars.

3.3 Memory

The in-vehicle storage market has great potential. From the current mainstream solutions for in-vehicle storage, the overall trend is that the number of storage chips used, the capacity of a single chip, and the value of a single chip are all increasing. A McKinsey report predicts the overall output value of in-vehicle storage, and it is expected that by 2020, the overall output value of in-vehicle storage will reach US$2.832 billion, of which DRAM and NAND will account for 51% and 36% respectively.

We have calculated the penetration of smart driving in China and the DRAM space. Based on the 2020 passenger car sales in China of 27.7 million, the smart car penetration rate of 50%, and the DRAM capacity of 38GB per vehicle, the in-vehicle DRAM space in China alone is expected to reach 527 million GB.

3.4 PCB

The use of automotive PCBs has increased. Printed Circuit Board (PCB) refers to a printed board that forms point-to-point connections and printed components on a general substrate according to a predetermined design. Its main function is to connect various electronic components to a predetermined circuit and play a role in relay transmission. PCB is an important carrier in the chip packaging process. In the era of intelligent networking, its use will increase with the increase in the number of automotive chips.

The automotive PCB market size increased from USD 2.8 billion in 2009 to USD 4.9 billion in 2015, with a compound annual growth rate of 9.78%. It is expected that the market size will be close to USD 6 billion in 2018. According to public data, automotive PCBs with less than 16 layers account for about 73%; HDI, FPC, and IC substrates account for 9.56%, 14.57%, and 2.38% respectively, accounting for about 27% in total. Therefore, multi-layer boards are still the main demand for automotive electronics. Many PCB manufacturers have also laid out in the automotive field. For example, Shanghai Electric Co., Ltd. and Dongshan Precision have already entered Tesla's supply chain and entered the intelligent networked car market.