Yole Développement, a research and strategy consulting firm, and its partner System Plus Consulting have been analyzing the semiconductor industry for several years, particularly in the field of automotive applications. OnMarch 23, they published a study entitled “Automotive industry: a massive AI-powered transformation is ongoing” which reveals the decisive role of the automotive market for the development of semiconductor innovations.

Yole and System Plus Consulting analysts collaborate on a daily basis to gain a global understanding of the industry, questioning technical trends and market developments. Using a dedicated methodology and expertise in semiconductor technologies applied to the automotive sector, both companies publish numerous analyses, including technology reports and market studies.

The autonomous car market

Road mobility is a US$10 trillion market. The CASE (connected, autonomous, shared and electrified) megatrends have taken the industry by storm. Cars sold today are adopting new imaging technology equipment to support new functionality, and in particular support the transition from advanced driver assistance systems (ADAS) to autonomous driving (AD).

This technology relies on cameras, radar, LiDAR, edge and cloud computing. Yole automotive analysts forecast ADAS to AD technology to grow at a CAGR of 12% for the period 2021-2027, reaching, at the electronic module level, US$25 billion by 2027.

Pierrick Boulay, Senior Technology & Market Analyst, Lighting and Display at Yole, explains:

“Today, only 12.3% of the world’s 1,300 million cars on the road are equipped with ADAS technology. This will increase to 49% of the estimated 1,800 million cars on the road by 2030. This is due to the huge increase in technology penetration in cars produced each year. From 21 percent five years ago, ADAS is now installed on 65 percent of all vehicle production in 2022, and that number will reach 86 percent by 2027.”

The commercialization of autonomous cars will still take time, however the study predicts that spending on electronic modules for ADAS to AD technology will double in the next five years. So far, ADAS technology is focused on radar, which is expected to reach US$9.3 billion, followed by camera modules comprising an image sensor, optics and camera housing will account for US$8.9 billion.

According to the report, imaging using standard CIS is being pushed to the limits of its perceptual capabilities. “Brute force” are the words used by some companies to describe the high computational power strategy being developed by AI companies to solve AD challenges through “camera only” approaches.

Other players are taking a different approach, including through the introduction of LiDAR and thermal imaging solutions, and even SWIR imaging solutions in the near future.

Adrien Sanchez, technology and market analyst, IT at Yole, comments:

“AD technology is not yet fully here for automotive consumers, while fleets of robotic cabs using a different set of hardware and software are starting to develop in a number of cities. At a figure of US$100,000 per car, there are already a few thousand robotic cars doing the job. AI technology combined with extremely powerful chipsets will also require different sensor modalities; these are the main factors driving autonomy.”

A hotly contested market

Semiconductor companies such as Mobileye, Nvidia and Qualcomm have carved out a niche for themselves in the automated driving systems market, providing the chips needed to process data from cameras, radar and LiDAR. At the same time, consumer market players such as Sony, Huawei, Alibaba, Baidu… have partnered with established OEMs to build a new car and provide the autonomous/automated driving technology.

The recent partnership between Honda and Sony is an example, with Honda manufacturing the car and Sony providing expertise in the development and application of imaging, sensing, telecommunication, networking and entertainment technologies.

In the past two years, many new OEMs have emerged in the automotive sector and are building electric cars around semiconductors and AI software for AD. Among them, XPeng, Nio, Link&Co, IM motors, and Arcfox, are integrating an average of six ADAS cameras and four display cameras into their cars plus two in-cabin cameras, five radars, and up to five LiDAR…

Translated from Voitures autonomes : une transformation massive alimentée par l’IA est en cours