
Embracing the "Great Era of Optics": Let Embodied Intelligence See the World

With the arrival of the era of embodied intelligence, the demand for optical technology has significantly increased in various fields such as smartphones, smart vehicles, drones, robots, and wearable devices. Major optical manufacturers like SUNNY OPTICAL, AAC TECH, and Q TECH are expected to benefit from this trend. The upgrade of optical specifications for smartphones, the popularization of cameras and LiDAR for smart driving, the expansion of drone vision systems, and the demand for robotic vision hardware will all drive a significant increase in the gross margin and net profit margin of related manufacturers
Abstract
Optics, the eye of embodied intelligence. In the era of embodied intelligence, hardware devices in multiple scenarios such as smartphones, smart cars, drones, robots, and wearables are ushering in a wave of AI innovation. Vision is the core perceptual link between embodied intelligence and the world. We believe that leading optical manufacturers such as SUNNY OPTICAL, AAC TECH, and Q TECH are expected to significantly benefit from the growth brought by the optical demands of embodied intelligence across multiple scenarios.
Smartphones: Optical specification upgrades are underway. Smartphones are the core AI hardware. As users' demand for high-performance imaging specifications becomes more pronounced, optical manufacturers are increasingly refining their explorations in areas such as periscope, optical image stabilization, large sensors, and miniaturization. Looking ahead to 2025, the trend of smartphone optical upgrades, represented by periscopes, is expected to become more evident. We anticipate that this will significantly boost the actual utilization rates and ASP of optical manufacturers, thereby providing considerable growth elasticity to their gross and net profit margins.
Automobiles: Equal rights in intelligent driving, cameras and LiDAR welcome a big year. 2025 is the year of equal rights in intelligent driving, with multiple automakers significantly accelerating progress in intelligent driving. Cameras, as a core component of the perception system essential for intelligent driving, are expected to see a substantial increase in the number of models equipped with intelligent driving solutions, as well as a significant rise in the average number of cameras per vehicle. With the decreasing costs of LiDAR, it has also become a reality for a larger number of models to be equipped with LiDAR.
Drones: In the low-altitude era, vision comes first. The visual system of drones includes both obstacle avoidance systems and imaging systems. As the number of drones (UAVs) continues to expand, and manned autonomous flying vehicles and cargo autonomous flying vehicles (AAVs) gradually enter operation, we expect the quantity and specifications of visual perception systems and imaging systems to continue to rise.
Robots: Visual hardware brings brightness to industrial, consumer, and humanoid robots. The visual system of robots mainly consists of cameras, depth cameras, and LiDAR. In the future, as consumer robots, industrial robots, and humanoid robots become more widespread, we anticipate that the demand for optical components such as cameras, depth cameras, and LiDAR in robotic scenarios will increase exponentially.
Wearables: Not just good glasses, but also good cameras. With the iteration of AI glasses products, the demand for photo interaction from users and manufacturers is also increasing. Rokid Glasses, Thunderbird V3, and others are equipped with 12 million pixel cameras and emphasize the "good camera" function. We expect that as AI glasses become more widely adopted, the demand for camera products in this scenario will see significant growth.
Broad prospects, recreating an optical track. We quantitatively estimate the future demand scale for optics across multiple scenarios: based on the judgment and assumptions regarding the upgrade of smartphone optical specifications and the growth in shipments of optical components for vehicles, drones, robots, and wearables, we estimate that the scale of optical components represented by cameras, depth cameras, and LiDAR is expected to exceed 300 billion RMB in 2024 and rise to over 700 billion RMB by around 2030, roughly equivalent to recreating an optical track Recommendations: SUNNY OPTICAL (2382.HK), AAC TECH (2018.HK), Q TECH (1478.HK), etc. Focus: Largan Precision, OFILM, SUTENG, GoerTek, Crystal Optoelectronics, Lianchuang Electronics, etc.
Risk Warning: The speed of mobile optical upgrades may not meet expectations, the penetration rates of smart driving, drones, robots, and camera glasses may rise or decline slower than expected, competition in the sector may exceed expectations, and calculation errors may exceed expectations.
Main Text
1. Optical Components: The World is Right Before My Eyes
1.1 Mobile Phones: Optical Specification Upgrades in Progress
Currently, we see that on one hand, mobile phone users' demand for high-performance imaging specifications is increasingly prominent, while on the other hand, optical manufacturers are making significant advancements in product exploration in areas such as periscope, stabilization, large format, and miniaturization. Looking ahead to 2025, we anticipate that the penetration rates of periscope, optical stabilization, and large format specifications in mobile terminals are expected to significantly increase.
In H2 2024, we expect that in the new product launches from Apple/Xiaomi/VIVO/OPPO/Honor/Huawei, periscope telephoto will become a commonly equipped feature. Apple's periscope telephoto extends from the iPhone 15 Pro Max down to the iPhone 16 Pro/Pro Max models. Xiaomi's periscope telephoto extends from the Xiaomi 14 Ultra down to the Xiaomi 15 Pro model. Oppo, Vivo, and Huawei have also equipped their flagship models in the second half of the year with periscope telephoto.
Looking ahead, we expect the penetration rate of high-end modules represented by periscopes to continue to rise significantly, driven by:
-
Optical upgrades are a perceptible and visible change for consumers. Upgrades such as periscopes significantly improve shooting effects and have strong demand.
-
In H2 2023, global mobile phone shipments are expected to rebound from the bottom, easing the inventory destocking phase of components.
-
Since Huawei's strong return, the catfish effect of optical upgrades has become prominent, with HOV competing to showcase their strengths.
-
The cost of mobile phone storage is expected to start declining, freeing up cost space for subsequent optical upgrades.
Q TECH expects that in China, the shipment volume of periscope modules is expected to increase from 46kk in 2024 to 65kk in 2025 and 80kk in 2026; the adoption of OIS optical stabilization and VA variable aperture is also expected to rise rapidly. From a global perspective, Q TECH anticipates that the shipment volume of periscope camera modules is expected to increase from 136kk in 2024 to 165kk in 2025 We are optimistic about the optical upgrade opportunities brought by the iteration and penetration of products such as periscope telephoto, multispectral modules, optical image stabilization, variable aperture, and glass-plastic hybrid lenses. We expect that optical upgrades are likely to significantly boost suppliers' utilization rates and ASP, thereby providing considerable growth elasticity to their gross profit margins and net profit margins.
1.2 Smart Vehicles: Equal Rights in Intelligent Driving, Cameras and LiDAR Welcome a Year of Adoption
2025 is the year of equal rights in intelligent driving, with multiple automakers significantly accelerating progress in the intelligent driving sector and continuously extending intelligent driving functions to mid-range models.
Taking BYD as an example, on February 10, BYD held a strategic release conference for intelligence, launching its universal intelligent driving strategy. Its "Tian Shen Zhi Yan" technology matrix will cover a full range of different models. Among them, the "Tian Shen Zhi Yan" C platform adopts a pure vision 5R12V12U solution, equipped with 12 onboard camera modules, with the first batch of 21 models covering price ranges from 70,000 to 200,000. The "Tian Shen Zhi Yan" B platform is equipped with LiDAR, mainly installed on Denza and some BYD models. The "Tian Shen Zhi Yan" A platform is equipped with three LiDAR units, primarily installed on the Yangwang models.
As a core component of the perception system essential for intelligent driving, we expect that with the popularization of intelligent driving, not only will the number of models equipped with intelligent driving solutions increase significantly, but the average number of cameras per vehicle is also expected to rise substantially.
In the LiDAR segment, its detection distance, reliability, pedestrian discrimination ability, and nighttime detection capability complement cameras well, providing good blind spot coverage. Qiu Chunchao, CEO of Suoteng Juchuang, stated, "The MX product will achieve mass production of the first project at a cost of less than $200, and we will quickly lock the price around 1,000 yuan, leading the entire industry into the era of thousand-yuan machines." As the cost of LiDAR decreases, it has become a reality for more models to be equipped with LiDAR.
1.3 Drones: In the Low Altitude Era, Vision Comes First
In the low-altitude economic era, whether it is drones (UAVs), manned autonomous flying vehicles, or cargo autonomous flying vehicles (AAVs), the importance of visual systems is increasingly prominent. Taking drones as an example, the drone vision system includes both perception and obstacle avoidance systems and imaging systems.
-
Perception and Obstacle Avoidance System: DJI's drone products can be equipped with the Guidance visual sensing navigation system, which combines monocular and stereo vision systems with ultrasonic sensors to achieve multi-sensor fusion, providing strong perception reliability for the flying platform.
-
Imaging System: The DJI aerial drone Mavic 3 Pro series is equipped with a three-camera system consisting of a 20-megapixel Hasselblad main camera and dual telephoto lenses of 12 megapixels and 48 megapixels
In the era of low-altitude flight, as the number of drones (UAVs) continues to expand, and manned autonomous aircraft and cargo autonomous aircraft (AAVs) gradually enter operation, we expect the quantity and specifications of visual perception systems and imaging systems to continue to rise.
1.4 Robots: Visual Hardware Brings Light to Robots
For consumer robots, industrial robots, and the evolving humanoid robots, the visual system is the core component of perception. The visual systems configured by different robot manufacturers vary slightly but mainly consist of cameras, depth cameras, and LiDAR.
-
Taking the Unitree G1 and H1 humanoid robots as examples, they are equipped with a visual hardware system composed of "depth camera + 3D LiDAR."
-
Taking the Expedition A2 from Zhiyuan Robotics as an example, it is equipped with a fusion perception system of "6 high-definition cameras + 360-degree LiDAR."
-
Taking the Cloud Ginger 2.0 from Dafab Robotics as an example, it is equipped with a multi-sensor fusion perception system of "RGB monocular camera + 3D depth camera/Single-point TOF camera + LiDAR."
In the future, with the popularization of consumer robots and industrial robots, as well as the accelerated progress of humanoid robots in product iteration and scene establishment, we expect the demand for optical components such as cameras, depth cameras, and LiDAR in robotic scenarios to increase exponentially.
1.5 Wearables: Not Just Good Glasses, But Also Good Cameras
With the iteration of AI glasses products, the demand for photo interaction from users and manufacturers is also increasing.
-
Rokid Glasses highlights "good camera" — featuring a 12-megapixel high-definition camera that supports vertical shooting and captures exciting moments quickly.
-
The Thunderbird V3 shooting glasses set the slogan as "The universe at this moment, all in sight" — its 12-megapixel wide-angle camera promotes "First-person Vlog, if you can see it, you can shoot it."
We expect that with the iterative popularization of AI glasses, the demand for camera products will significantly increase in this scenario.
2. Vast Space, Recreating an Optical Track
We predict the potential space of the optical components track based on five major scenarios: mobile phones, automobiles, drones, robots, and glasses.
- Mobile Phones
-
Core Driver: Upgrades in optical specifications lead to an increase in the unit value of mobile phone camera modules.
-
Calculation Assumptions:
➔ Mobile Phone Shipment Volume: We assume that the global mobile phone shipment volume will maintain a year-on-year growth rate of 3.0-4.5% from 2025 to 2026, and around 2% year-on-year growth from 2027 to 2030.
➔ Camera Module Value: With the iterative upgrade trends such as periscope, large aperture, variable aperture, and lens-side glass-plastic hybrid, we assume that the ASP growth rate of Android mobile phone camera modules will be around 10% from 2025 to 2026, and around 3% from 2027 to 2030; we assume that Apple's CCM will maintain an ASP growth rate of around 3% in most years.
- Smart Vehicles
-
Core Driver: The popularization of intelligent driving leads to an increase in 1) the shipment volume of in-vehicle cameras and 2) LiDAR.
-
Calculation Assumptions:
➔ Smart Vehicle Delivery Volume: According to data from ZhiShi Consulting, the global number of vehicles equipped with intelligent driving functions is expected to be around 43 million in 2024, potentially exceeding 81 million by 2030.
➔ Camera Installation Volume:
★ Number of Cameras per Vehicle: According to statistics from ChaoDian Think Tank, the global shipment volume of in-vehicle camera modules is approximately 260 million/340 million in 2023/2024, and we estimate the average number of cameras per vehicle to be 6.6/7.9. We assume that by 2030, the average number of cameras per vehicle will approach 12.
★ ASP of In-Vehicle Camera Modules: We assume that the price of in-vehicle camera modules will be in the range of 200-300 yuan in 2024, and we assume a single-digit annual price decline.
➔ LiDAR Installation Volume:
★ LiDAR Installation Rate: We assume that as the cost of LiDAR decreases, the installation rate of LiDAR in smart vehicles is expected to increase.
★ LiDAR ASP: We assume that as the cost of LiDAR decreases, the average price of LiDAR is expected to drop to the thousand yuan price range.
- Drones
-
Core Driver: The growth in drone sales leads to an increase in the shipment volume of in-vehicle camera modules.
-
Calculation Assumptions:
➔ Drone Delivery Volume: The State Council Information Office disclosed that the shipment volume of civilian drones in China in 2023 is 3.17 million units. We assume that the civilian type accounts for 75% of the total drone volume and that China's shipment volume accounts for 70% of the global total.
➔ Single Unit Value of Drone Vision Systems: We assume that with the improvement of aerial photography specifications, the total value of the perception obstacle avoidance system + imaging system per unit will increase from hundreds of yuan to the thousand yuan level 4) Various Types of Robots
-
Core Drivers: Significant growth in shipments driven by industrial scenarios, consumer service scenarios, humanoid robots, etc.
-
Calculation Assumptions:
➔ Robot Shipments: According to IFR statistics, in 2023, approximately 211,000 consumer service robots (including specialized and medical robots) are expected to be shipped, while industrial robot shipments are expected to be around 541,000 units. According to GGII statistics and forecasts, humanoid robot shipments are expected to reach 12,000 units in 2024 and exceed 50,000 units by 2026.
➔ Single Machine Value of Robot Vision Systems: As the robot sector is still in its early stages, the single machine value of optical components varies significantly. We estimate it ranges from thousands to tens of thousands of yuan. We expect that as the robot sector continues to iterate, the specifications of optical components will improve, and the average price will reflect economies of scale.
- Smart Glasses
-
Core Drivers: Growth in AI glasses shipments leads to an increase in camera shipments.
-
Calculation Assumptions:
➔ Glasses Shipments: According to Wellsenn XR, global AI smart glasses shipments are expected to be around 2 million units in 2024, reaching 80 million units by 2030.
➔ Single Machine Value of Glasses Cameras: According to Wellsenn XR, AI glasses manufacturers with a relatively mature supply chain can control the value of the camera module on the glasses to around 10 USD.
Based on our rough predictions, considering the upgrade of mobile optical specifications and the growth in shipments of optical components for vehicles, drones, robots, wearables, etc., we estimate that the scale of optical components represented by cameras, depth cameras, and LiDAR is expected to exceed 300 billion RMB in 2024 and rise to over 700 billion RMB by 2030, effectively creating a new optical sector.
We are optimistic about the vertical specification upgrades and horizontal scenario expansions in the optical sector, as well as the sustainable growth that will result from these developments.
Recommendations: SUNNY OPTICAL (2382.HK), AAC TECH (2018.HK), Q TECH (1478.HK), etc.
Focus: Largan Precision, OFILM, SUTENG JUCHUANG, GoerTek, Crystal Optoelectronics, Lianchuang Electronics, etc.
Risk Warning
The speed of mobile optical upgrades may fall short of expectations. If the upgrade trends of periscope, large aperture, miniaturization, and other mobile optical components do not meet expectations, the ASP and GPM of related companies may be affected.
The risk of penetration rates for smart driving, drones, robots, and camera glasses not meeting expectations. If the promotion of new scenarios for related smart hardware does not meet expectations, the optical segment may also be impacted Risks of competition in the optical sector exceeding expectations. If competition in the optical sector intensifies, it may affect the performance of various companies.
Calculation errors exceeding expectations. The future scale estimation process of the sector includes our assumptions about the shipment volume of various intelligent hardware and the value of components. If the deviation of the assumptions from the actual situation is large, the calculation error may exceed expectations.
Authors of this article: Xia Jun, Liu Ling, Source: GS Overseas Research, Original title: "The Great Era of Optics: Let Intelligent Hardware See the World | Guosheng Overseas"
Risk warning and disclaimer
The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investment based on this is at one's own risk