Is it really hard to find rivals for Mobileye in the fields of ADAS and autonomous driving?

 
Currently, Mobileye's Road Network Collection Management™ (REM™) in China's travel sector has covered Beijing, Shanghai, Guangzhou, Hong Kong, Macau, and Taipei, covering commercial vehicles, freight vehicles, logistics vehicles, buses & long-distance buses, shared cars & taxis, rental cars, RVs & leisure vehicles, law enforcement vehicles, etc. Last year, the two smart travel-related projects in China were jointly conducted by Volkswagen and Beijing Public Transport. Volkswagen has been able to achieve L2+ level autonomous driving by equipping its cars with front cameras and Mobileye's Road Book™ technology.

Responsibility-Sensitive Safety Model (RSS)

Considering that driverless cars will share the roads with traditional human-driven cars in the next few decades, the Responsibility Sensitive Safety (RSS) model advocated by Mobileye will be a very important link.

In 2017, Intel and Mobileye proposed an open, transparent, and verifiable formal model, the Responsibility Sensitive Safety (RSS) model, hoping to enable self-driving cars to judge their own safety status by establishing mathematical formulas. The model explains how to promote the development of the autonomous driving industry by regulating accident faults and vehicle safety. The Responsibility Sensitive Safety model is a catalyst that can promote cross-industry discussions among industry organizations, automakers, and regulators. One of the key issues is that human judgment is based on legal, safety, and cultural considerations, while the RSS model seems to focus only on the legal level.

RSS materializes the concept of human safe driving into a verifiable model with logically verifiable rules and defines appropriate response behaviors. The model is technology neutral, which enables the entire industry to use it as a starting point to unify the understanding of safe autonomous driving. When RSS can become an open standard accepted and used by the entire industry, companies only need to share data related to RSS, and there will be no problem of leaking core intellectual property rights.

Intel and Moibleye believe that betting on the field of autonomous driving is to promote industry consensus - it is absolutely necessary to standardize the rules of judgment, responsibility and fault in order to achieve huge benefits to society. RSS is not a model to avoid responsibility, but an innovation model, the purpose of which is to enable autonomous vehicles to operate according to the highest safety standards. For autonomous driving to succeed, the entire ecosystem must be mobilized to allow scientific experiments to enter the real market.

Currently, based on RSS, Intel and Mobileye are not only cooperating with OEMs and industry suppliers, but also actively communicating with regulatory agencies in the United States and China. In July this year, Baidu announced that it would deploy the RSS model in its Apollo open source project and Apollo Pilot commercial project. The two parties will also jointly verify the RSS model based on the unique driving style and road conditions in the Chinese market, and update and improve the RSS model based on new discoveries in the cooperation.

 
In October this year, Mobileye and Volkswagen announced the joint launch of the first autonomous ride-hailing service (also known as Mobility as a Service, MaaS) in Israel. The focus of this project is that Israel will launch the world's first government-involved autonomous driving MaaS project. The Israeli government has promised to set up a special committee to begin discussing regulatory barriers and research related matters. This not only achieves technological innovation, but also promotes regulatory and institutional innovation.

Aftermarket ADAS solution: Mobileye Shield+
In order to further strengthen the safety technology management of operating buses and effectively curb and reduce road transport safety accidents caused by insufficient inherent safety performance of buses, the Ministry of Transport issued the "Safety Technical Conditions for Operating Buses" (JT/T1094-2016) standard in March 2017. The document requires that from April 1, 2018, all operating buses with a length of more than 9 meters need to be equipped with LDW (lane departure warning system) and AEBS (automatic emergency braking system) forward collision warning functions. From April 1, 2019, newly produced vehicles should have other functions of AEBS.

This means that commercial vehicle AEBS will usher in a market explosion, with an estimated annual market size of more than 450,000 sets. On this basis, Mobileye has released its latest aftermarket ADAS solution - Mobileye Shield+. Mobileye Shield+ is mainly installed on large vehicles such as buses and trucks. This ADAS solution mainly uses 6 sensors to detect lane departure, intelligent high beam control, front collision warning, and pedestrian and cyclist warnings. In the past year, Mobileye's anti-collision technology and its products such as Mobileye Shield+ for large urban vehicles have been recognized and installed by many manufacturers and government departments.

The artificial vision intelligent camera array used by the Mobileye Shield+ system is like a third eye for the driver, monitoring potential dangers in blind spots in real time, and giving the driver enough time to react through different sound and light alarms to avoid collision accidents. By using artificial vision intelligent cameras to monitor and analyze the real-time dynamic driving road environment, and automatically ignoring stationary objects and pedestrians in safe areas, the false alarm rate is reduced.

In early September 2017, Mobileye officially announced a strategic partnership with Chinese bus manufacturer Yutong. Mobileye will install Mobileye's ADAS system on every new energy bus produced thereafter.

Mobileye mainly expands the market around the EyeQ series chips and chip-based ADAS (advanced driver assistance systems), strengthening Intel's autonomous driving capabilities through sensing and vision perspectives.

Competition intensifies in the ADAS and autonomous driving fields

Mobileye believes that only when advanced technology enters the market to benefit mankind can technology truly realize its value.

In recent years, Mobileye has promoted the development of autonomous driving and assisted driving through multi-field and all-round cooperation. At the same time, as automakers and technology giants scramble to bring driverless cars to the mass market, tens of billions of dollars are pouring into the field of autonomous driving technology.

With the increase of competitors in this field, market competition has become more intense.

There are overlaps with traditional semiconductor manufacturers
such as Mobileye and NXP, but Intel is not involved in relatively basic microcontrollers. 

In the field of autonomous driving cameras, there are at least ten chip companies involved, such as NXP, Texas Instruments, Renesas Electronics and Infineon. Mobileye's key advantage lies in its early entry into the field; secondly, the company has proposed an optimized computing architecture specifically for computer vision computing, which is similar to the concept of "accelerator devices" and can improve the performance and efficiency of camera lens computing while maintaining a very low energy consumption performance.

With the increase of competitors in this field, market competition has become more intense.

 
Nvidia
Intel and its long-time rival Nvidia have yet to decide who is the winner in the autonomous driving platform, but the two companies are approaching the market in different directions. Nvidia is mainly targeting autonomous driving above Level 3, while Intel has plans from Level 1 to Level 5. 

Nvidia has been active in the autonomous driving business in recent years. First, it launched the DGX SuperPOD, the world's 22nd fastest supercomputer for autonomous driving technology calculations, and then announced the deepening of cooperation with the Volvo Group to develop autonomous driving systems. It can be seen that Nvidia is actively building a one-stop shopping ecosystem for its own autonomous driving solutions, and is gradually forming a global open autonomous driving processor competition with Intel's Mobileye, competing with Tesla's closed autonomous driving hardware development.

So far, many automakers have cooperated with NVIDIA and adopted NVIDIA's Drive self-driving platform. In addition to Volvo Group, Japan's Toyota, Germany's Audi, and Mercedes-Benz's parent company Daimler are also automakers that cooperate with NVIDIA's self-driving platform; Bosch, Continental, and ZF, and other major auto parts suppliers, also have a cooperative relationship with NVIDIA; Baidu Apollo's self-driving ecosystem is also a partner of NVIDIA. Uber's self-driving system also uses NVIDIA's self-driving hardware, but NVIDIA has always emphasized that Uber does not use its Drive platform.

Now NVIDIA has launched the DGX SuperPOD self-driving supercomputer, providing a more complete one-stop shopping product portfolio for self-driving development customers: NVIDIA previously launched Drive IX, which mainly supports artificial intelligence (AI) in-vehicle infotainment systems; the Drive Pegasus platform provides the computing requirements for in-vehicle full self-driving functions; Drive Constellation can provide a wide range of self-driving system simulations to improve the safety and efficiency of self-driving systems. DGX SuperPOD can provide self-driving technology developers with more realistic neural network training challenges.

The DGX SuperPOD contains 96 Nvidia DGX-2H supercomputers with a total of 1,536 Nvidia V100 GPUs, and is interconnected with Nvidia NVSwitch and Mellanox network fabrics, providing 9.4 petaflops of computing power to help automakers, startups and researchers train neural networks for their self-driving applications. Nvidia bought Mellanox for $6.9 billion in March this year.