Research on the Automotive Optics Industry

I. Automotive Lenses: The Eyes of Autonomous Driving – A High-Growth, Long-Term Opportunity in a Thriving Market

1. High Growth Potential: The automotive lens market is projected to exceed RMB 33 billion by 2025, with a Compound Annual Growth Rate (CAGR) of 37%.
   Automotive cameras are indispensable sensors for autonomous driving systems. These systems typically consist of three layers: Perception, Decision-Making, and Execution. Sensors used in the perception layer include visual sensors like cameras, LiDAR, millimeter-wave radar, ultrasonic radar, as well as speed and acceleration sensors. Compared to other sensors, cameras possess strong obstacle recognition capabilities, making them essential components for autonomous driving.

   Classification of Automotive Cameras:

   • By Function: Imaging cameras (for passive safety, storing/displaying images) and Perception/ADAS cameras (for active safety, requiring precise image capture).

   • By Position: Front-view, Side-view, Surround-view, Rear-view, and In-cabin cameras.

     ① Front-View Cameras: Enable various ADAS functions (collision warning, lane departure warning, etc.). They bear significant responsibility and require the highest specifications. Front-view setups can be monocular, binocular, or trifocal. Monocular cameras, developed earlier, are technologically mature and cost-effective for mass production but are limited by their fixed focal length, struggling to balance wide-angle (short-range) coverage with telephoto (long-range) detection. Binocular and multi-camera systems partially overcome these limitations. For example, Tesla's trifocal camera includes: a primary camera (covering most driving scenarios, max detection range ~150m), a wide-angle camera (150° field of view, capturing traffic lights, obstacles, and nearby objects ideal for city streets/low speeds), and a telephoto camera (clearly capturing distant objects for highway driving, max detection range ~250m).
     ② Side-View Cameras: Monitor blind spots ahead or behind the vehicle.
     ③ Surround-View Cameras: Use wide-angle lenses installed around the vehicle to create a stitched 360° bird's-eye view, enabling panoramic parking assistance. With algorithms, they can also perceive lane markings.
     ④ Rear-View Cameras: Use wide-angle lenses primarily for reversing assistance.
     ⑤ In-Cabin Cameras: Monitor driver status for drowsiness alerts.

   Advanced Autonomy Drives Quantity & Value Growth: Higher levels of autonomous driving (AD) push demand for more and better cameras. AD levels range from L0 to L5; currently, mainstream adoption lies between L2 and L3. L2 features include backup monitoring, panoramic park assist, blind spot detection, adaptive cruise control, forward collision warning, intelligent speed assistance, lane departure warning, pedestrian detection, and traffic sign recognition – typically requiring 3-13 cameras. L4/L5 AD systems, still under development, generally require 13+ cameras. Furthermore, unlike traditional rear-view cameras capturing mostly static images, ADAS cameras must capture clear moving objects at high vehicle speeds, demanding higher specifications and unit prices. Additionally, as AD computing power increases, higher resolution automotive cameras will be needed.

   Market Acceleration Post-2017: 2017 marked the dawn of widespread advanced driver-assistance systems (ADAS). Rapid adoption of L2-level ADAS is expected following its breakout year in 2021.

   • Tesla's Catalytic Role: As a leader in EVs and ADAS, Tesla launched its first electric sedan, the Model S (with L2 ADAS and 8 cameras), in 2012 at $$50kUSD.Highcost,limitedproductioncapacity,andinfrastructureconstraintskeptitniche.In2017,Tesladisruptedthemarketwiththeentry−levelModel3(alsoL2ADAS,8cameras)pricedat50k USD. High cost, limited production capacity, and infrastructure constraints kept it niche. In 2017, Tesla disrupted the market with the entry-level Model 3 (also L2 ADAS, 8 cameras) priced at 50kUSD.Highcost,limitedproductioncapacity,andinfrastructureconstraintskeptitniche.In2017,Tesladisruptedthemarketwiththeentry−levelModel3(alsoL2ADAS,8cameras)pricedat35k USD, helping it capture an 8% market share. Subsequently, benefits from its Shanghai Gigafactory and economies of scale reduced the Model 3 price to ~¥240k RMB by August 2021.
   • Industry Response: Motivated by Tesla's "catfish effect," other NEV manufacturers and legacy automakers are rapidly adopting ADAS systems and associated hardware. Given OEM model cycles exceed 3 years, L2 penetration surged significantly in 2021. Future ADAS penetration is expected to follow a steep S-curve similar to smartphone adoption, indicating rapid growth.

   Market Projections: We forecast an average of 8 cameras per vehicle by 2025, leading to an automotive camera market exceeding RMB 160 billion and an automotive lens market surpassing RMB 33 billion (CAGR >37%).

   • Baseline Data: Global automotive camera shipments reached 165 million units in 2020, growing at a CAGR of 30% over the past decade, averaging 2.1 cameras per vehicle.
   • Projection Details: Based on subsequent calculations, we estimate 8 cameras per vehicle by 2025, resulting in 800 million camera shipments (CAGR=37%). Assuming stable pricing, this translates to a lens market size of ~RMB 33.6 billion. Considering that camera modules typically cost 5x the lens price, the automotive camera module market is projected to reach RMB 168.4 billion by 2025.