Seize the "commanding heights" of autonomous driving, and the traditional Tier1 giant "throws money into the market"

Traditional auto parts suppliers are looking to seize the commanding heights of autonomous driving technology and the market.

This week, Bosch announced the acquisition of British self-driving startup Five for an undisclosed amount (the latter was valued at around $200-300 million after the latest funding round). Upon completion of the transaction, the latter will become part of the Cross-Domain Computing Solutions Division, strengthening the development capabilities of autonomous driving systems.

Bosch said the acquisition is still awaiting approval from regulators, particularly antitrust authorities. The deal will probably take two months to complete.

Founded in 2016 and employing about 140 people, Five focuses on cloud-based software development and testing platforms for software development and testing of autonomous driving systems. For example, the system is verified by a subset of real road data acquisition and simulation.

"Five's engineering capabilities, especially the mindset of our employees and agile development methods, have the opportunity to help us move one step closer to our goal of safe autonomous driving." Mathias Pillin, President of Bosch Cross-Domain Computing Solutions, said.

This year, Bosch won the software platform cooperation development order of Cariad, a software subsidiary of the Volkswagen Group, which means that Tier1, which traditionally uses hardware as its main business, needs to increase investment in software capabilities.

At the same time, the downstream market demand for high-end automatic driving has emerged, including some car companies have begun to enter the L3, L4 level automatic driving system mass production preparations. For the traditional Tier1 with ADAS assisted driving as the main market, it is necessary to improve the product line layout as soon as possible.

What types of businesses are popular

Unlike other companies focused on mass production and on-the-ground operation of L4 autonomous driving, Five saw different market opportunities at the beginning of its inception and focused on creating development tools and platforms to help test and verify the safety of autonomous driving systems.

At present, Five's main business is mainly divided into two parts, one is the automatic driving development and safety verification platform (the main service is to create simulation scenarios, evaluate system performance, and optimize software algorithms), and the other is related software stack customization services.

The above-mentioned customized development services are mainly for several scenarios: 1. Development of ALKS (single-lane automatic driving) based on specific ODDs; 2. Highway automatic driving system; 3. Urban autonomous driving.

Ceo Stan Boland, a former co-founder and CEO of Era Inc. and Element 14 Inc., was later acquired by Nvidia and Broadcoms for $1 billion. He was also the CEO of Acorn Computer Company and a member of the ARM Board of Directors.

In fact, Five's positioning is precisely because Stan Boland felt the "unrealistic hype and expectations" of the industry during the early road test and system development of L4 level autonomous driving.

After raising $41 million in Series B funding in 2020, Stan Boland was determined to transform his business: to productize highly integrated, cloud-based workflows and deliver the platform to downstream customers (e.g., OEMs, Tier1, and other L4 autonomous driving companies) to develop, test, and validate autonomous driving systems.

"Focusing on B2B, rather than B2C, means we have more customers and start to realize what the industry really needs – a convergence of capabilities between tech companies and traditional automakers." Stan Boland said.

In fact, the business model behind this is also very clear. No matter how the downstream development, platforms and tools are long-term demand, and at the same time, B2B's cash flow is definitely earlier than B2C landing.

In Stan Boland's view, the L4 level autonomous driving system of many companies can already be proven in a very good environment, but it will take a while to extend this method to a very complex urban environment, in the changeable weather and light conditions.

At the same time, the industry can clearly expect that L4 and even L5 driverless systems can cope with the landing operation of non-public roads and non-human vehicle mixed traffic conditions at low speeds and specific routes.

In the private consumer market, car companies should not only seek to "display" autonomous driving functions in the high-end market, but also need to promote assisted driving technology in a large area in the low-end market. However, from the perspective of actual development, a unified platform, framework and test verification process can greatly reduce unnecessary costs.

Stan Boland said that in the case of the ALKS regulations, it starts with single lane keeping and adaptive cruise control under speed limit conditions, and then gradually advances the landing by increasing the speed and increasing the complexity of the scene.

"This is a more realistic goal and path." In Stan Boland's view, both known and unknown environmental factors pose a huge challenge to the security of the system, and all sensors, modules and subsystems are exposed to limitations and vulnerability risks, which, together with regulatory factors, constitute an extremely complex problem.

This has also directly driven Five's transformation: "We realized that the right thing to do is to focus on the key parts. ”

For example, in early 2021, Five and Cognata (Simulation Tool Platform) announced the joint launch of a modular, cloud-based, end-to-end development automatic lane keeping system (ALKS) test platform, which will help car companies and Tier1 significantly reduce development and verification costs.

The layout point in time appears

From the perspective of industrial layout, the time point (how much advance) is a strategic issue that the traditional auto Tier1 giant must consider. Early input means that short-term output is often lower than expected; late input, there is a risk of being crowded out of the market by new entrants.

Earlier this year, after the failed acquisition of Veoneer, Magna instead announced the acquisition of optimus Ride, a self-driving company, thereby strengthening the company's market competitiveness in driver assistance systems, "With the gradual landing of autonomous driving, we see the urgency of rapidly introducing external technical capabilities, because this is the core need of customers in the future." ”

Last October, another auto parts giant, ZF, announced its stake in Oxbotica, a British self-driving car software startup, and set out to jointly develop driverless shuttles for people and things, while enhancing its own self-driving technology.

ZF also took a stake in South Korean assistive/autonomous driving sensing software providerStradvision this year to expand its portfolio of autonomous driving sensing software, including key components for autonomous driving systems for driverless feeders.

In fact, as early as 2018 ZF has launched the miniature driverless shuttle bus e. Go Mover, and plans to achieve mass production in 2019, with an initial plan for annual production of five figures. The company also expects market demand to reach about 1 million units over the next five to seven years.

It is clear that ZF wants to quickly integrate external capabilities and accelerate opportunities for new business expansion through investment participations and acquisitions. "We don't want to wait, we want to provide turnkey solutions." Torsten Gollewski, Head of Automated Transport Solutions at ZF, said.

And for Bosch, the same is true.

In January 2021, Bosch's Cross-Domain Computing Solutions Division was officially established to develop automotive hardware and software for specific applications, including driver assistance systems, infotainment, connected services, and autonomous driving. This is a major strategic restructuring as the world's auto parts leader, involving nearly 17,000 engineers worldwide.

And Bosch is already feeling the pressure on the smart driving track.

According to the monitoring data of The Institute of Intelligent Vehicles of Gaogong, at present, in the front-loading intelligent driving domain controller and system solution market, Chinese suppliers such as Desay SV, Fortetech, Huawei, Neusoft Ruichi, Millima Zhixing, Moshi Intelligent, Jimu Intelligent, Intelligent Driving Technology and other Chinese suppliers and foundries like Flextronics have emerged.

In the large-power domain controller market such as Nvidia, Desay SV has led the traditional foreign Tier1. In contrast, bosch's previously launched DASy assisted driving domain controller can only meet the requirements of ordinary L2 level assisted driving. At the same time, Bosch's traditional advantages of the 1V1R/1V3R track have also been "besieged" by Chinese local suppliers.

Just earlier this month, Volkswagen Group China announced that it was adopting a more Focused China sourcing strategy, inviting Chinese technology companies to participate in the Group's global procurement for the first time to further promote the strategic transformation with software as the core.

Although Bosch had also received software projects from the Volkswagen Group before this, it is clear that these joint venture brands and overseas local markets, which were originally almost monopolized by foreign Tier1, are also beginning to seek new opportunities for the reconstruction of the supply chain system.

In recent years, Bosch has also begun to focus on strategic upgrading at the software level, and at the same time began to participate in the leading enterprises in China's local market segments.

In September last year, Boyuan Capital officially announced its official establishment, and announced the landing of its first RMB fund, "Bosch China Growth Fund Phase I". This is an investment platform set up by the Bosch Group for the Chinese market, focusing on the fields of automotive and mobility services.

Prior to this, Bosch had invested in a number of local automotive technology companies in China, including Momenta, Hesai, Yishi Technology, Mainline Technology, and Chelian Tianxia. Among them, Bosch has cooperated with Chelian Tianxia and has mass-produced cockpit domain controller projects in the Chinese market.

At the same time, during this year's CES exhibition, Bosch announced that it will invest $4.5 billion per year (of which about $3.4 billion will be invested in mobility) to enhance the company's software competitiveness and once again emphasize the transformation strategy to a software-driven company.

"Investment will be an important means for the Bosch Group to expand its business in China," said Chen Yudong, President of Bosch China, adding that the long-term development strategy of further expanding local R&D and innovation advantages remains unchanged.

On January 12, 2022, Black Sesame Intelligence, a Chinese autonomous driving chip company, announced that it has received a strategic investment from Boyuan Capital, and the two sides will deepen comprehensive cooperation in the field of autonomous driving and work together to create intelligent driving solutions.

On February 18, Shanghai Geometry Partners Intelligent Driving Co., Ltd., a total solution provider for autonomous driving (a software and hardware integration system based on machine perception and learning for L2-L4), announced the completion of a new round of strategic financing, which was exclusively invested by Boyuan Capital.

For Chinese local suppliers, with the help of large-scale mass production in the Chinese market, they are at the forefront of the global market in terms of cost control ability and performance verification.

This is a major opportunity for industrial restructuring in the era of intelligence.