LiDAR industry current situation：
This year has seen numerous variables, and the LiDAR industry remains quite urgent. With an increasing number of car models entering the market and scheduled for release, the industry is experiencing a supply-demand imbalance.
fig. 1 LiDAR Road Testing
LiDAR currently achieves mass production mainly through MEMS and scanning mirror solutions. Represented by companies like Robosense, Hesai, Innovusion, and Neuvition, these technologies are the core solutions for LiDAR mass production. Some secondary companies like Zvision are also gradually emerging, but MEMS and scanning mirror solutions will remain the optimal choices for large-scale production in the next five years.
Flash, FMCW, and OPA LiDAR Technologies:
Flash LiDAR Solution: Currently, the maturity level of Flash is not particularly high. One reason is that Flash requires a large array, including a large array of Single-Photon Avalanche Diode (SPAD) receivers. Sony has started mass-producing VGA solutions this year, but it is still in the sampling phase. VGA resolution is insufficient for detecting distances of 50 meters or even slightly longer. Additionally, the large-array VCSEL chips used for transmission also pose challenges in terms of manufacturability and yield. Therefore, Flash solution faces considerable challenges in achieving mass production in the next one to two years.
fig. 2 Flash LiDAR
FMCW LiDAR Solution
In both China and internationally, the next-generation LiDAR is shifting towards FMCW solutions. Although the fastest samples may be introduced in 2025, led mainly by companies like Intel and AEVA in the United States, China players like Robosense, Hesai, and Neuvition have also scheduled some engineering samples for 2025 or 2026. However, true mass production may be later due to the long time required for the coherent silicon photonics chips and the related industrial chain from the light source to the receiving end.
fig 3 AEVA FMCW LiDAR
OPA LiDAR Solution
OPA will likely be even later as it only replaces the current MEMS or scanning mirror scanning methods and may also match the FMCW solution. Overall, OPA’s rollout is contingent upon the maturity of the FMCW technology. Currently, OPA has issues with high power consumption, sideband effects, overall flexibility, and the maturity of the chip, making it less viable for use in main LiDAR applications compared to MEMS and scanning mirrors at this stage.
fig 4 Quanergy OPA LiDAR
In summary, regarding scanning methods, MEMS and scanning mirrors, particularly MEMS, will be the future mainstream even if FMCW achieves first-generation mass production. This year may start with some difficulties, but as more car models are launched, especially with an increasing number of releases and upcoming models from automakers like BMW, Mercedes-Benz, Volkswagen, and Volvo, the LiDAR industry will enter a period of rapid development over the next three years.
It will also be the first round of industry consolidation, where the focus lies on the ability to achieve large-scale production, vehicle quality control, cost control, and delivery capabilities. In the second phase, the emphasis will shift to technology iteration, which will be the main direction of attention beyond the next three years. In conclusion, LiDAR is currently on the verge of mass production, particularly in the second half of the year. Although the volume may be slightly lower than initially projected for this year, the industry’s certainty is still high. Over the next few years, the technology will converge, and the industry will undergo consolidation.
Q&A:
1. What is the price difference between different categories of LiDAR technical routes? And the trend of cost reduction and price reduction in the future?
LiDAR systems operating at 905nm and 1550nm have different price ranges. The 905nm LiDAR systems are generally priced at around $600-800 USD, and they may not be profitable at these price levels. On the other hand, 1550nm LiDAR systems are more expensive, starting at around $2000 USD and even reaching $3000-4000 USD for companies like Luminar. The pricing largely depends on the scale of production and the maturity of the industry chain. In the domestic market, except for NIO, most LiDAR manufacturers currently use the 905nm wavelength due to its mature and cost-effective industry chain.
Regarding the future cost reduction trend, it is expected that in the next 2-3 years, and possibly until 2025, the prices of 905nm LiDAR systems may not decrease significantly and could stabilize at around $500 USD. The main focus for the first half of this year and next year will be on achieving rapid mass production rather than immediate profitability. The goal is to ensure large-scale production capability and stay ahead of competitors. Cost considerations are likely to become more prominent in the second half of next year and beyond when the industry focuses on controlling costs and achieving economies of scale.
By 2024, companies may intensify efforts to control costs, and at that time, the market will witness significant changes in the LiDAR industry. However, it is not expected that the prices of 905nm LiDAR systems will drop significantly below $400 USD.
2. What is the current progress of FMCW and OPA LiDAR? Which one is more optimistic about the industry at present?
FMCW technology has made significant progress and is now close to mass production. It relies heavily on silicon photonics technology, using narrow linewidth light sources and rotating mirrors for MEMS scanning at the receiving end. To achieve coherent reception and enable mass production, existing silicon photonics technology plays a vital role.
In the industry, the United States, led by Intel and Aeva, is at the forefront, and they are expected to introduce FMCW solutions as early as 2025.
For pure chip companies, excelling in FMCW presents challenges. Customized transmitter components and the need for precise alignment with receiver indicators make module manufacturing difficult. Understanding LiDAR technology, coherent reception, specification, and software algorithms are crucial.
While companies like Intel and Aeva have a strong position in the FMCW market, others may face difficulties without prior LiDAR experience. Collaboration with established LiDAR companies might be the best way forward for newer entrants, allowing them to leverage expertise and ensure successful FMCW development.
3. Once these two routes are mass-produced, the existing LiDAR technical routes may be replaced. What is the specific rhythm judgment?
By 2025, LiDAR products that have reached a level suitable for installation in vehicles will undergo fixed-point testing. However, true mass production and deployment on a broader scale are estimated to occur around 2027-2028. During this period, the LiDAR industry will likely undergo grading, meaning different technical routes will coexist, and one route may not entirely replace the other. Time-of-flight (TOF) solutions, utilizing Dtof technology with 905 or 1550 wavelengths, will continue to dominate the industry for the next five years or even longer.
Despite the introduction of Frequency Modulated Continuous Wave (FMCW) technology in the early stages, it is expected that it will be matched with well-established scanning solutions like MEMS or rotating mirrors. The gradual shift towards FMCW will likely occur over time, and a complete replacement of existing TOF solutions might not be immediate.
Overall, the LiDAR industry will experience a period of coexistence and gradual adoption of FMCW, but TOF-based solutions will remain prevalent for the foreseeable future. The continuous development and integration of advanced technologies will shape the LiDAR market as it evolves to meet the demands of the automotive industry and other sectors.
4. OPA may come out later than FMCW, and the difficulty may be higher.
Yes, that’s correct. Optical Phased Array (OPA) technology may indeed be introduced later than Frequency Modulated Continuous Wave (FMCW) LiDAR, and it poses higher technical challenges. OPA is associated with significant loss, side lobes, and complex algorithms, making its implementation and chip production more difficult compared to other LiDAR technologies.
In the short term, using OPA alone for scanning might not provide substantial advantages. However, the value of OPA is expected to increase when combined with FMCW to form a unified and comprehensive solution. By integrating OPA with FMCW, their complementary strengths can be leveraged to improve. Including the overall performance and capabilities of the LiDAR system.