Software-defined vehicle Research Report 2021: Architecture Trends and Industry Panorama

[ad_1]

New York, Aug. 11, 2021 (GLOBE NEWSWIRE) — Reportlinker.com announces the release of the report “Software-defined vehicle Research Report 2021: Architecture Trends and Industry Panorama” – https://www.reportlinker.com/p06128353/?utm_source=GNW
;
(3) The application middleware and development framework, including functional software, SOA, etc.;
(4) The application software layer, including smart cockpit HMI, ADAS/AD algorithms, connectivity algorithms, cloud platforms, etc.

The business model of smart car software is in the form of “IP + solutions + services”. Tier 1 software suppliers charge in following modes: (1) One-time R&D expenses and software packages, such as ADAS/AD algorithm packages; (2) Software licensing fees per car, royalties (by a certain proportion of car shipments and unit price); (3) One-time R&D expenses and license packages per car.

Take the software IP licensing fees as an example. Regardless of the highly complex AD software, we estimate that the software IP licensing fees for a single car is at least RMB2,000-3,000. As the functions of smart cars become more complex, the software IP licensing fees per car will continue to rise.

At the same time, OEMs are also vigorously expanding their internal software R&D teams to reduce the cost of software outsourcing. In the next step, OEMs‘ software R&D will still focus on software that can directly create value for consumers, such as cockpit HMI, autonomous driving, etc. Of course, OEMs can also cooperate with independent software vendors in R&D through the decoupling of software and hardware, like BMW and ArcherMind Technology team up, and ThunderSoft and Human Horizons collaborate. However, for general software (such as surround view splicing, voice, DMS) and common platform-based software (such as OS kernels, virtual machines, HD maps, cloud platforms, etc.), OEMs still give priority to outsourcing.

In general, as software becomes more and more complex, the life cycle value (ASP) of software for a single car may be as high as tens of thousands of yuan, which makes software a main part of the vehicle BOM cost.

With the continuous evolution of software-defined vehicles, the business model of the entire automotive industry has changed accordingly, from new car manufacturing and marketing model which lasts for a long time to a larger-scale software × ownership model. Automakers will charge customers for software license and OTA updates to complete the closed-loop business model.

For example, Tesla has boosted the price of its FSD (Full Self-Driving) Beta launched at the end of 2020 by US$2,000 to US$10,000, and it will further raise the price of the package for L3/L4 autonomous driving to US$14,000. Tesla is trying to offer customers a way to subscribe and pay a monthly fee instead of US$10,000 up-front for their premium driver assistance features in order to expand the potential subscriber base.

On July 16, 2021, Tesla started offering a monthly subscription for its FSD package for US$199 per month in the United States. Tesla owners can cancel their monthly FSD subscription at any time. Customers who previously bought Tesla’s Enhanced Autopilot package (EAP) can subscribe to FSD for a lower price of US$99 a month but may require the HW3.0 upgrade. In other words, only Teslas that have the FSD computer hardware 3.0 (HW3) or above plus either Basic or Enhanced Autopilot configurations are eligible to subscribe to FSD; other owners can purchase a hardware upgrade for US$1,500 to make their vehicles FSD-ready.

Once Tesla fulfills the transformation of its business model, all existing Tesla owners may become subscribers of the FSD package. Assuming 10 million Tesla owners subscribe to the FSD package with a monthly price of US$100, a subscription fee of US$12 billion will be incurred each year and Tesla’s software gross margin will be as high as 70-80%. The software revenue from existing car owners is expected to be very stable and lucrative, posing a solid moat for Tesla.
A Trilogy of Software-defined Vehicle Transformation of OEMs

In the short term, the system kernel and middle layer are the key R&D directions. In the long run, SOA will bring about changes in business models. To complete the software-defined vehicle transformation, OEMs must at least achieve:

1. Vehicle EEA upgrade. The hardware architecture develops from distributed ECU to centralized domains, and further upgrades to centralized + regional controllers. As for the communication architecture, the automotive network backbone is upgraded from LIN/CAN bus to Ethernet;

2. Linux, QNX and other RTOS only provide kernels. On this basis, OEMs realize hardware abstraction, form a middle layer operating system that supports application development, define developer interaction logic, and build application layers, namely the so-called self-developed operating systems of OEMs, similar to Tesla.OS, VW.OS, Daimler MB.OS, BMW-OS and Toyota Arene. At the same time, more and more OEMs have followed suit, such as SAIC Z-One SOA, Lixiang Li-OS and Volvo Cars.OS. The ultimate goal of OEMs is to open up vehicle programming to all enterprises by simplifying the development of vehicle software and increasing the frequency of updates, so as to master the ecological resources of developers;

3. With further making use of the huge number of users to build a developer ecosystem, the profit engine of automakers has transferred from “hardware manufacturing” to “software development”. For example, Tesla continues to boost the price of FSD software while continuously reducing the price of vehicle hardware in order to promote the rapid development of the automobile company dominated by software revenue.

In the short term, most OEMs are still in the stage of upgrading their hardware architecture. At present, only Tesla and Volkswagen have completed the development, construction and large-scale applications of customized OS kernels. The decoupling of automotive hardware and software is also in the early stages of development. Now, OEMs have focused on basic software (system kernels, AP Autosar, middle layers, etc.).

From a long-term perspective, SOA (Service-Oriented Architecture) will reconstruct the automotive ecology. The automotive industry is likely to replicate the software model of “basic hardware, middle-level operating systems and upper-level applications” for PCs and smartphones. Meanwhile, smart car middleware giants will emerge. Upper-level APP developers need not pay attention to the underlying hardware architecture, but should focus on application development instead.

By building operating systems and SOA platforms by themselves or with suppliers, automakers have introduced a large number of algorithm suppliers, ecological partners, etc. to form a developer ecosystem. In the future, automakers can provide users with software services covering the full life cycle. Under this background, OEMs have laid out SOA software architecture development. In the next 2-3 years, SOA mass production will reach its peak, and consumers will enjoy richer smart driving experience.

The automotive SOA software platform is similar to Apple iOS and Google Android in the smart phone field. It is not only a generalized software architecture, but also an ecological platform for developers. In the smart phone field, Apple iOS and Google Android basically monopolize developer resources, each of them holds more than 20 million developer resources worldwide. Likewise, the automotive SOA software platform may gradually head toward an oligopolistic market.

At the same time, the powerful ecosystem of Apple and Android developers has become one of the fatal technologies in the cont
ext of the Sino-US trade war. Especially after the US government placed Huawei on its Entity List, Google stopped offering Google Mobile Services to Huawei, which affected the operation of Huawei’s mobile phones. In this case, Huawei had to launch HMS services and HarmonyOS. The underlying automotive software platform will also threaten the safety of the industrial strategy, and it has become a national strategy. It is urgent to build a standardized platform for automotive basic software.

Under the guidance of the Ministry of Industry and Information Technology of China, Neusoft Reach and China Association of Automobile Manufacturers have jointly planned and initiated China Automotive Basic Software Ecological Committee (AUTOSEMO). With more than 20 member units, AUTOSEMO will share…

[ad_2]

Read More:Software-defined vehicle Research Report 2021: Architecture Trends and Industry Panorama