Source: Zero Power Industry Research Institute  20^th Nov 2025

Introduction: The Dual Challenges of Technological Breakthroughs and Industrialization Bottlenecks

As a core sensing component for autonomous driving, robotics, smart cities and other fields, the technological iteration and industrialization process of LiDAR have always been accompanied by in-depth contradictions among stability, cost and large-scale application. Early mechanical LiDAR failed to meet automotive-grade requirements due to its complex structure and poor reliability. While solid-state technology is regarded as the ultimate solution, it still faces constant pressure for optimization in aspects such as environmental adaptability and mass production yield. At the same time, the persistently high cost of LiDAR has directly restricted its penetration in mid-to-low-end vehicle models and consumer-grade scenarios.

I. Industry Status Quo: Dual Drivers of Technological Iteration and Market Expansion

1. Diversification of Technical Routes: Solid-State and Chip-Scale Solutions Become Mainstream Directions

According to the analysis in 2025–2030 China LiDAR Industry In-Depth Research and Development Status & Trend Report released by Zhongyan Puhua Industry Research Institute, the current LiDAR technical routes exhibit distinct characteristics of "mechanical models fading out, semi-solid-state solutions taking the lead, and solid-state technologies emerging". Mechanical LiDAR, plagued by short service life and high costs, has gradually withdrawn from the automotive sector and shifted its focus to niche markets such as surveying and mapping, as well as security and protection. Semi-solid-state solutions (including MEMS micromirrors and rotating mirror types) have become the preferred choice of major automakers due to their balanced advantages in technological maturity and cost-effectiveness. Take MEMS as an example: by leveraging microminiaturized vibrating mirrors to steer laser beams, it has successfully reduced the thickness of LiDAR units to a level suitable for concealed installation, thus resolving the challenge of vehicle body integration. In contrast, fully solid-state Flash LiDAR eliminates mechanical moving parts by adopting VCSEL arrays and SPAD receivers. The chip-scale solution integrates transmitting, receiving, and processing modules into a single ASIC chip, cutting power consumption by over 60% compared with traditional mechanical counterparts, and has emerged as the core track for future technological competition.

The divergence between the 905nm and 1550nm wavelength routes persists. 905nm lasers are based on the mature GaAs material system, boasting significant cost advantages. However, their emission power is constrained by eye-safety standards, resulting in a shorter detection range. On the other hand, the 1550nm wavelength is safer for human eyes as it is absorbed by the cornea and lens, enabling longer-distance detection. Yet, it relies on costly InGaAs materials, with costs several times higher than those of 905nm alternatives. Although 1550nm holds a foothold in the high-end market, 905nm continues to dominate the mid-to-low-end market through technological upgrades (such as multi-beam stacking and high-power pulses), and the market landscape is unlikely to be overturned in the short term.

2. Market Size: Explosive Growth in the Automotive Sector and Expansion into Diversified Scenarios

The global LiDAR market is experiencing a period of rapid growth. In 2022, the global market size of LiDAR solutions reached RMB 12 billion. By 2024, the size of China’s LiDAR market exceeded RMB 13.9 billion, with the automotive sector accounting for over 60% of the total. This growth is mainly driven by the large-scale commercialization of L3-level autonomous driving. According to forecasts by Zhongyan Puhua, the global automotive LiDAR market size will surpass USD 6 billion in 2025, with the penetration rate of L2+ vehicles exceeding 20%, driving the domestic passenger car installation volume to break through 13 million units.

In terms of application scenarios, LiDAR is accelerating its penetration from autonomous driving to intelligent transportation, robotics, surveying and mapping, and other fields. For instance, in smart cities, LiDAR constructs real-time urban perception networks using 3D point cloud data to enable functions such as traffic flow monitoring and parking space guidance. In the industrial sector, AGV robots equipped with LiDAR achieve high-precision navigation, driving the flexible upgrading of intelligent manufacturing. Statistics show that in 2024, the demand for LiDAR in the service robot (cleaning and delivery) sector surged by 120% year-on-year. The rise of the humanoid robot market has further spawned customized demand for miniaturized, low-power-consumption LiDAR units.

3. Competitive Landscape: Chinese Manufacturers Dominate the Global Market

The global LiDAR market has formed a top-tier "four-horse race" pattern. Four major manufacturers—Huawei, Hesai Technology, RoboSense, and Innovusion—account for over 95% of the global automotive LiDAR market share. Among them, Hesai Technology leads the market with its ATX series of automotive-grade LiDAR. Huawei has built ecological barriers through its integrated "hardware + perception algorithms + data services" solutions. RoboSense, focusing on digital technologies, has launched its EM platform LiDAR with over 500 lines to meet the requirements of L4-level autonomous driving.

The rise of Chinese manufacturers is attributed to three core advantages:

Independent Technological Control: The localization rate of core components exceeds 90%, benefiting from the maturity of local supply chains such as laser devices from Focuslight Technologies and SPAD detectors from Mira Photonics.

Cost Leadership: Through architectural innovation and mass production scale effects, Hesai Technology has reduced the unit price of LiDAR from tens of thousands of US dollars to below USD 200.

Policy Dividends: China has incorporated LiDAR into the mandatory national standard Safety Requirements for Combined Advanced Driver Assistance Systems of Intelligent Connected Vehicles, stipulating that L3-level autonomous driving vehicles must be equipped with LiDAR. Local governments are promoting industrial development through subsidies and scenario opening-up initiatives.

II. Development Trends: Future Landscape of Technological Integration and Scenario Expansion

1. Technological Trends: Intelligent Algorithms and Multi-Sensor Fusion

According to forecasts in the 2025–2030 China LiDAR Industry In-Depth Research and Development Status & Trend Report by Zhongyan Puhua Industry Research Institute, the core competitiveness of LiDAR in the future will center on intelligent algorithms and multi-sensor fusion capabilities. Enterprises are achieving raw-layer fusion of data from LiDAR, cameras, and millimeter-wave radars through BEV bird's-eye view networks and Transformer architectures, enhancing target recognition accuracy in complex scenarios. For example, in scenarios where LiDAR performance is compromised (such as rainy or foggy weather), 4D millimeter-wave radars can supplement data to ensure full-time safety. In the autonomous driving sector, the implementation of L3-level regulations has driven LiDAR to become a standard configuration. It is projected that the penetration rate of L3-level vehicles in China will exceed 30% by 2026, fueling sustained growth in market demand.

In addition, the point cloud density and detection range of LiDAR will continue to improve. Evolving from traditional 16-channel to 128-channel and even 300-channel+ configurations, point cloud density will increase by more than 10 times, strengthening the ability to identify small obstacles. The detection range of mass-produced products will extend from 150 meters to over 300 meters, enabling early decision-making in high-speed scenarios. The popularization of FMCW (Frequency-Modulated Continuous Wave) technology will further resolve ambient light interference issues, while directly acquiring velocity data to replace traditional ToF radars.

2. Market Trends: From High-End Configuration to Mass Popularization

The automotive LiDAR market is undergoing concurrent transformations of "volume growth coupled with price decline" and "performance leap". On one hand, the surge in demand from automakers has driven annual production capacity to exceed one million units, with costs dropping to the USD 500–1000 range (2023 level). Chinese manufacturers aim to lower costs to below USD 200 by 2025, promoting the popularization of L2+ vehicles. On the other hand, solid-state technology is gradually maturing, and fully solid-state Flash LiDAR has entered the technology optimization phase. It is estimated that solid-state solutions will account for over 70% of the market by 2030, with costs falling to USD 100, becoming a standard sensor for mid-to-high-end vehicle models.

In the robotics sector, LiDAR applications exhibit the characteristics of "diversification + customization". Service robots leverage LiDAR for SLAM navigation and dynamic obstacle avoidance. The consumer-grade market has relatively low cost sensitivity, driving demand for high-end products. The humanoid robot market imposes stringent requirements on the size, weight, and real-time performance of LiDAR, prompting enterprises to develop specialized solutions. For example, models such as UBtech Walker S and Xiaomi CyberOne use LiDAR to construct 3D environmental models, enabling gesture recognition and spatial interaction.

3. Industrial Ecosystem: Vertical Integration and Global Layout

The LiDAR industry chain is accelerating vertical integration, with leading enterprises building barriers through full-chain layout covering "chips – modules – systems – applications". The localization rate of upstream core components (e.g., VCSEL chips, SPAD receivers) continues to rise, yet high-end ASIC processing chips still rely on imports. Midstream enterprises are enhancing competitiveness through self-built production lines (e.g., Hesai Technology's German factory) and ecosystem collaborations (e.g., joint development between RoboSense and NVIDIA). Downstream application scenarios are continuously expanding, extending from automotive to consumer electronics, drones, and other fields.

In terms of globalization, Chinese enterprises are expanding overseas markets through localized production and technology licensing. For instance, RoboSense has partnered with Stellantis to develop automotive LiDAR, and Huawei is building an intelligent driving ecosystem with global automakers via its "Huawei Inside" model. At the policy level, China has incorporated LiDAR into mandatory standards for intelligent connected vehicles, while European and American countries are driving technological standardization through ISO 26262 functional safety certification (ASIL-B/D), laying the regulatory foundation for global market competition.

The LiDAR industry is in a critical transition phase from the "technology verification period" to the "large-scale commercialization period". Technologically, the integration of solid-state, chip-scale solutions and intelligent algorithms will drive continuous performance improvements and cost reductions. In the market, the mainstream adoption in the automotive sector and the diversification of robotics and smart scenarios will jointly form the "dual engines" of industry growth. For the industrial ecosystem, vertical integration and global layout will reshape the competitive landscape, and Chinese manufacturers are poised to dominate the global market leveraging independent technological control and cost advantages.

Nevertheless, the industry still faces challenges such as insufficient technological maturity, supply chain security risks, and intensifying international competition. In the future, enterprises that can accurately grasp technological trends, deeply understand scenario requirements, and flexibly adjust strategies will stand out amid this transformation and lead a new chapter in the era of intelligent perception.