Cameras are now standard equipment for automobiles, with 124 million image sensors shipped in 2018. Automotive camera modules have reached $3bn and are expected to grow at an 11% CAGR, reaching $5.7bn by 2024, announces Yole Développement (Yole) in its latest imaging report, Imaging for Automotive.
Over the past five years, viewing applications have been at the core of market growth – with rearview, surround-view, and black box becoming ubiquitous. ADAS cameras, which currently represent 40% of the business, will provide additional growth for years to come thanks to growing adoption rates.
As an example, ZF, one of the largest tier one suppliers of automotive systems, last year released its fourth Generation S-Cam with two solutions, one with a mono camera and the other with a triple camera set-up.
Both companies, Yole and System Plus Consulting are working together to propose a deep understanding of the imaging technologies and their automotive applications. Their analyzes include an AI-powered vision computing to reveal additional momentum to numerous automotive applications and imaging technologies.
From a market and technology perspectives, analysts are offering today a wide collection of reports to describe the imaging industry, innovative technologies and emerging market segments. In addition to the Imaging for Automotive report announced today, they present both reverse engineering & costing reports: ZF S-Cam 4 – Forward Automotive Mono and Tri Camera for Advanced Driver Assistance Systems.
Yole and System Plus Consulting, both part of Yole Group of Companies combine their expertise to point out technical breakthroughs and market opportunities. What is the status of imaging technologies within the historical automotive industry? What will be the tomorrow’s industry? How will the car manufacturers and Tiers-1 manage this evolution? Analysts invite you to discover the future.
Formerly considered as a secondary market for imaging companies, the automotive industry is now central to every key player’s strategy: firstly because revenue and volume have reached a significant level even when compared to mobile, and secondly because of AI far-reaching impact, which powers ADAS.
“As the first full-scale edge computing application, imaging for automotive has positioned itself as a key “center of gravity” for imaging technologies during the transition from imaging to sensing applications”, said Pierre Cambou, Imaging Activity Leader at Yole. “New functionalities and novel technologies are today transforming the entire automotive industry.”
Yesterday, automotive applications used to be straightforward: rearview cameras and forward ADAS cameras, for the most part. Today, the situation has dramatically changed and the technological consequences may prove dramatic as well.
Yohann Tschudi, PhD. Technology & Market Analyst at Yole, explained: “For example, on the viewing side of the market, surround-view cameras crave expansion in lower-end markets, which should change the number of ISPs. Then there are the mirror replacement cameras, also called “e-mirrors” requiring anti flickering and HDR. That is why technical performance should far transcend the current viewing cameras.”
The fourth Generation S-Cam proposed by ZF, with its two solutions, one with a mono camera and the other with a triple camera set-up features the Omnivision CMOS image sensor.
“This technical choice made by ZF, demonstrates the shift in the procurement strategies of ZF and Intel Mobileye,” commented Audrey Lahrach, Cost Analyst from System Plus Consulting. “The use of the latest Mobileye EyeQ4 vision processor allows to combine the information from three cameras to improve security thank to a computing capacity multiplied per 10. All in one of the smallest and lightest products in its category.”
On the ADAS side of the market, the in-cabin application is being highly scrutinized by many CIS players, mostly pushed by regulations. Moving forward, market acceptance will be interesting to monitor.
The big question for reaching Level 3 autonomy is the technology selected for ADAS surround cameras. Data overload is obviously around the corner, and brute-force computing approaches, i.e. using a 180Tops ASIC (such as the FSD from Tesla) will not be possible or preferred for every OEM .
A new technology paradigm is brewing, but typical of automotive technology’s slow rate of introduction, it could take three to five years to manifest. The collection of imaging reports is available on i-micronews.com, imaging reports section.