Editor's note: " New variable " is a column launched by Automobile Heart to share insights from front-line practitioners of smart cars. From the perspective of witnesses, we will take you to foresee the key variables in the development of smart cars.
Author of this article: Zhou Yanwu, a senior expert in the industry, and a special author of Auto Heart.
2020 is an epoch-making year for the lidar industry.
Velodyne and Luminar have successively entered the US stock market, and Innoviz, Aeva and Ouster are on the way.
, the five major lidar companies, currently the highest market value of Luminar exceeds 10 billion US dollars, and the lowest valuation of Innoviz also reaches 1.4 billion US dollars.
Behind these companies are popular in the capital market, major automakers have chosen to carry lidar on mass-produced cars.
Honda and Toyota have determined to use lidar on their L3-class autonomous driving models; manufacturers such as Mercedes-Benz, Volvo, BMW, NIO and Xiaopeng are also preparing to use lidar on mass-produced cars in 2021.
It can be said that the golden age of lidar is coming.
This article will analyze the technical routes of the five major lidars and take you to see where the popular LiDAR chickens are heading.
The content of this article will include:
- Velodyne Core technology analysis: MLA
- Luminar: The highest power brings the highest performance
- Innoviz: MEMS brings the lowest cost
- Aeva: stick to FMCW
- Ouster: The technical route similar to Flash
1. Lidar basically constitutes
It is worth noting that Velodyne, Luminar, Innoviz, Aeva and Ouster, these 5 In terms of the main products of the front-loading vehicle market, all companies have abandoned the traditional bearing motor mechanical rotation scheme.
This is different from the current main products of , Huawei , Sagitar, Hesai, Radius, Yizhi and other companies in China.
Before analyzing each enterprise, we will give a brief introduction to the terms of lidar performance evaluation.
lidar is divided into pulse ToF type and continuous wave type according to the ranging principle.
Our common products are mostly pulse ToF type, which consists of four parts in hardware, namely:
- laser emission
- scanner
- reflected light reception
- data processing
continuous wave type lidar is divided into two types: phase modulation and frequency modulation, among which frequency modulation, FMCW, is more common.
and then specifically for the laser emission part, we usually divide it into three categories:
- EEL laser diode , usually 905 nanometers and 1550 nanometers. The materials include three types: silicon, GaAs (gallium arsenide) , and InP (indium phosphide) .
- VCSEL, vertical cavity surface emission type, usually in array form.
- fiber laser tube .
receiving part is usually divided into four categories:
- PIN diode, without any gain.
- APD, avalanche diode, has a certain degree of gain.
- SPAD, that is, single photon array, with ultra-high gain.
- MPPC or SiPM, similar to SPAD.
LiDAR as a sensor, its core indicator is signal-to-noise ratio . However, this is also an indicator that lidar companies never disclose.
2. Technical analysis of the five major lidar companies
(1) Velodyne's core technology: MLA
Velodyne is the originator of the field of vehicle lidar .
Velodyne started to develop solid-state lidar in 2015, released Velarray in 2017, and basically completed the design in 2020.
was designed by Hyundai Motor in the same year. It is said that GAC is also testing Velarray.
Velodyne has been working for the longest time in the field of lidar and has accumulated a lot of research and development results of mechanical lidar.
Previously, Velodyne said that the core technology on Velarray was not MEMS.
On Velarray, Velodyne Son microns the mechanical lidar. After the micron is reduced, it can adopt resonance scanning method, and still uses multiple laser emitters, so it is indeed not a MEMS.
Traditional mechanical lidar has three scanning methods:
The first is prism . The disadvantage is that it will introduce unnecessary size increase, and there will be wear of bearings or bushings, which will affect life over time; the advantage is that the number of lines can be very high. A typical representative is Huawei .
The second type is rotary mirror . The disadvantage is that it cannot fully utilize the time domain, has a certain amount of volume waste, low number of lines, and it is difficult to achieve high performance; the advantage is that it has a long life and high reliability. A typical representative is Valeo Scala.
The third type is MEMS galvanometer . The disadvantages are that FOV is limited, the reliability is doubtful, the signal-to-noise ratio is low, and the effective distance is short. The advantage is the low cost.
Velodyne has developed Resonant Mirror (resonance mirror) technology. The difference between this technology of
and the MEMS galvanometer is quite different.
resonance scanning does not have the disadvantages of the above three scanning mirrors, but it requires the laser radar to be reduced in microscope and needs to be combined with a concave mirror to form an arc shape. In addition, the cost will increase a lot.
Velodyne has applied for the patent for Resonant Mirror (resonance mirror) . The patent is shown in the figure above, where 163 and 164 are the core.
In the field of optical communication, optical resonance is the basic element, and optical resonance cavity is one of the three elements composed of laser amplifiers. It can be divided into open cavity and closed cavity according to the existence of the reflective surface. The function of the
resonance cavity is mainly used to enable the gain medium to achieve the inversion of the layout. It can be used as the optical amplifier (Gain amplifier) . Through the resonance cavity, the amplified signal can be collected to form an oscillator (oscillator) .
laser resonance cavity types are mainly divided into three categories:
The first type is parallel plane cavity .
consists of two parallel plane mirrors , which is optically called Fabry-Borot optical resonance cavity (Fabry-Perot resonator) , referred to as F-P cavity for short, and is mostly used in solid-state laser systems.
The second type is double cavity .
consists of two concave mirrors, one of which is a special and commonly used form is the confocal cavity (confocal) , which consists of two concave mirrors with the same radius of curvature. The distance between the two mirrors is equal to the radius of curvature. The two mirrors overlap with the focus, and the diffraction loss of the confocal cavity is small and easy to adjust.
The third type is flat cavity .
consists of a plane mirror and a concave mirror. One of the special and commonly used forms is the semi-confocal cavity, which is equivalent to half of the confocal cavity.
Judging from Velodyne's patent, the MLA array has a slight arc and should be combined with a concave reflector.
MLA Details, the picture shows the 8-channel module
Velodyne. The newly appointed CTO Mathew Rekow is from the field of optical communication and is very familiar with resonance cavity.
Velarray's main R&D work is Mathew Rekow.
One of his job is to miniaturize and modularize the lidar to improve the mass production efficiency of Velarray and reduce costs.
The difficulty of this work lies in the fact that the microscopic module also needs to ensure high performance, especially the laser diode emission requires relatively large current, and traditional power devices cannot meet the requirements.
To solve this problem, Velodyne began to cooperate with EPCh in 2016.
EPC is specialized in GaN power device technology. GaN is a wide bandgap semiconductor material. Field-effect transistors made of this material are more than 10 times faster than traditional transistor switching speeds.
Velarray uses GaN field effect tube, which is what Velodyne calls a custom ASIC.
Its body is actively small, only 2 to 4 mm2. While the size is reduced, its performance is also improved.
There is a simple formula in lidar. The Z-axis resolution of lidar depends on pulse width .
uses GaN field effect tube ASIC's Velarray pulse width up to 5 nanoseconds, which is the highest performance ASIC chip except SPAD.
Most solid-state lidars are generally 50-150 nanoseconds, and SPAD can easily achieve 1 nanosecond or even dozens of picoseconds.
Currently, Velarray's main product is 8-channel modules. High-performance products can use 4 to 16 modules, and low-performance only requires 1 module.
(2) Luminar: The highest power brings the highest performance
The easiest and most effective way to improve the performance of lidar is to increase the laser emission power.
While increasing power, you must consider the human eye safety of the product.
905 nanometer silicon photodetector, 550 nanometer InGaAs is 100,000 times safer, and you can safely increase the power of laser .
Luminar is characterized by using 1550 nanometers InGaAs.
The laser power used is 40 times that of traditional silicon photoelectric systems. Not only does it have a high signal-to-noise ratio, it reduces the pulse width to less than 20 nanoseconds, the pulse repetition frequency is less than 100MHz, and the duty cycle is less than 1%; at the same time, this also increases the effective distance.
In rainy, snowy and foggy days, the reflectivity of the object will be reduced, which will shorten the effective detection distance of the lidar, but increasing the power can solve this problem. That's what Luminar does.
Luminar emphasizes:
Even for objects with 10% reflectivity, the effective detection distance of their products can reach 200 meters.
Luminar also filed a patent for laser power amplification.
Its patent is a pulsed laser system with a secondary large-mode field erbium-doped fiber (EDFA) amplifier modulates a seed source laser into a pulse width below 20 nanoseconds, pulse repetition frequency below 100MHz, and duty cycle below 1%.
Luminar's patented core is a seed source laser, and the other is a bait-doped fiber amplifier.
The picture above shows the internal composition of Luminar's seed source laser
The picture above shows the internal composition of Luminar's amplifier
In terms of scanners, Luminar does not have much innovation, but still uses the traditional MEMS dual-axis galvanometer scanning.
Generally speaking, traditional MEMS lidar has low signal-to-noise ratio, but Luminar's power density is amazing, completely eliminating this disadvantage.
Due to the introduction of fiber lasers, the Luminar lidar is slightly larger.
In addition, the use of 1550 nanometer InGaAs lasers also keeps its product costs high.
Although Luminar has repeatedly emphasized that it has the ability to reduce costs, fiber lasers have been used for more than 20 years and have long lost the potential for performance mining.
Therefore, the industry has always had doubts about Luminar's cost control capabilities.
(3) Innoviz: The MEMS route brings the lowest cost
MEMS is the fastest implementation solution at present. Compared with mechanical lidar,
has three advantages:
First of all, MEMS micro galvanometer helps lidar get rid of bulky motors, polyprisms and other mechanical motion devices. The millimeter-level micro galvanometer greatly reduces the size of lidar and improves reliability.
Innoluce MEMS lidar schematic diagram
is secondly cost. The introduction of MEMS micro galvanometer can reduce the number of lasers and detectors, greatly reducing costs.
How many wire harnesses are needed for traditional mechanical lidars to realize, the corresponding number of transmitting modules and receiving modules is required.
uses a two-dimensional MEMS micro galvanometer, and only one laser light source is needed to reflect the laser beam through a MEMS micro galvanometer.
and the two use microsecond frequency to work together, and after receiving the detector, they achieve the purpose of 3D scanning of the target object.
Compared with mechanical lidar structures with multiple transmit/receive chipsets, MEMS lidar demand for lasers and detectors is significantly reduced.
From a cost perspective, N-line mechanical lidar requires N-group IC chipsets:
- transimpedance amplifier (TIA)
- low-noise amplifier (LNA)
- comparator (Comparator)
- analog-to-digital converter (ADC) , etc.
If imported laser (typical such as Excelitas LD) and detector (typical such as Hamamatsu PD) , the cost per line lidar is about 200 USD , domestic lasers such as commonly used Changchun optical machines can be lower.
MEMS can theoretically achieve its 1/16th cost.
finally has resolution, and the MEMS galvanometer can accurately control the deflection angle, rather than adjusting the motor speed like mechanical lidar.
For example:
Velarray has 2 million echo points per second.
and Velodyne's 128-line lidars only have 2.4 million, and Velarray is almost equivalent to 106-line mechanical lidar. What are the disadvantages of
MEMS? The disadvantages of
are that the signal-to-noise ratio is low, the effective distance is short, and the FOV is too narrow.
Because MEMS only uses one set of transmitting lasers and receiving devices, the signal optical power must be much lower than that of mechanical lidar.
At the same time, the light receiving aperture at the receiving end of the MEMS lidar is very small, much smaller than that of the mechanical lidar, and the peak light receiving power is proportional to the receiver aperture area, which leads to further power drop.
or above means that the signal-to-noise ratio is reduced, and the effective detection distance is shortened.
scanning system resolution is determined by the product of the mirror size and the maximum deflection angle.
The larger the mirror size, the smaller the deflection angle.
The larger the mirror size, the higher the resolution.
Finally, the cost and size of the MEMS galvanometer are also proportional.
Currently, the largest size of MEMS galvanometer is Mirrorcle, which can reach 7.5mm, and the price is as high as USD 1199 .
The MEMS micro galvanometer developed by Satin invested in Xijing Technology has a diameter of 5mm and has entered the mass production stage.
The MEMS micro galvanometer used in Hesai Technology PandarGT 3.0 was developed by the team.
There are two main solutions to the shortcomings of MEMS:
One is to use a laser with a 1550 nanometer emission wavelength to further increase the power using an erbium-doped amplifier in the optical fiber field. The human eye safety threshold of lasers in the
1550 nanoband is much higher than that of 905 nanometer lasers. Therefore, the power of the 1550-nanometer fiber laser can be greatly improved within the safe range. A typical example is Luminar.
's disadvantage is that 1550 nano lasers are extremely expensive.
and this is the category of the laser industry. Lidar manufacturers have far less technical accumulation than laser industry manufacturers in this regard, and it is almost impossible to reduce costs.
2 uses SPAD or SiPM to receive arrays, rather than traditional APD arrays, and the SPAD array efficiency is about 100,000 times higher than APD.
But the SPAD array is not very mature yet and the price is slightly higher.
(4) Aeva: insist on FMCW
lidar, traditional camera and mm wave radar have similarities. Traditional ToF lidar can be regarded as a 3D camera, but the resolution is generally very low.
Traditional cameras are 2D imaging, and lidar is 3D.
laser can also be regarded as an electromagnetic wave, which is also very close to millimeter wave radar.
FMCW lidar schematic
Early cars also used the direct emission and reflection of electromagnetic waves to measure distance. Later, it was found that this method has low signal-to-noise ratio and high power consumption, just like the current ToF lidar.
Later, it was found that continuous wave frequency modulation coherence detection (FMCW) has a high signal-to-noise ratio and low power consumption, but has a large signal processing operation.
With the improvement of chip computing power at present, this difficulty has been gradually overcome. Today, electromagnetic wave radars are all FMCW type.
In addition, early electromagnetic wave radars were also driven by motors, and later converted to printed plane antenna arrays instead of mechanical scanning.
People seem to be able to draw conclusions from the development of vehicle-mounted radar that the lidar will eventually be FMCW, and it also uses arrays instead of scanners.
ToF LiDAR has many interference factors or noise.
is the influence of sunlight, which is more sensitive to 1550 nanometer laser, and 905 nanometers is much better.
Second, the surface material and color of an object will also affect the laser absorption rate of different colors and materials is different. For example, the reflectivity of white and black is huge, and the reflectivity is closely related to the effective distance. The lower the reflectivity, the shorter the effective distance.
Generally, the test condition of 90% of the reflectivity is required to measure the effective distance of the lidar; if the reflectivity is 10%, in extreme cases, the effective distance may be shortened by 50%.
The number of black object reflected point clouds is low and may not be sensed at long distances.
FMCW LiDAR uses phase interference beat frequency method to measure, and these noises no longer exist.
For FMCW lidar, the signal-to-noise ratio is proportional to the total number of emitted photons, rather than the peak laser power.
Because the FMCW lidar has more than 10 times higher sensitivity, its average transmit power can be 100 times lower than the pulsed ToF lidar, which means low power consumption and higher levels of eye safety.
FMCW The photonic circuit of the lidar mixes a portion of the emitted coherent laser with the received light.
This provides a unique "unlock key" that can effectively prevent any environmental radiation or other lidar interference.
FMCW The light source of the lidar needs to modulate the frequency of the optical carrier in different forms according to the measurement purpose. Currently, commonly used include triangular wave form, sawtooth wave form and sinusoidal form. The frequency of the transmitted signal of
periodically changes around the optical carrier frequency fc with time t. Each period T is called the signal repetition time, and the frequency range (f1-f2) is called the modulation bandwidth B.
can easily demodulate the Doppler frequency of the target reflected signal using a triangular wave shape, thereby achieving simultaneous ranging and speed measurement.
The frequency modulation form of sawtooth wave shape is often used when the Doppler frequency shift introduced with the relative speed of the detection target can be negligible, and the relatively maximum detection distance can be achieved.
sinusoidal-shaped frequency modulation signal is relatively convenient, but the demodulation method is complex, and its accuracy is slightly worse than the modulation form of high frequency modulation linearity.
Generally, triangular waves can measure the speed of the target like FMCW millimeter wave radar.
Today's FMCW millimeter wave radar is very simple, the main chips are transceivers and processors, and the benefit it brings is that it is easy to integrate and chip. This also means small size and low cost.
However, the maturity of FMCW millimeter wave radar has lasted for nearly 10 years.
Today's FMCW lidar technology can be said to be very mature. Both laser modulation, reception and data processing are in the infancy and are far from comparable to ToF lidar.
, especially laser modulation, is extremely difficult and there are only a few companies engaged in related research.
According to the relationship between the tuning device and the laser, the current methods of implementing laser light carrier frequency modulation can be divided into internal modulation technology and external modulation technology.
internal modulation technology refers to a modulation technology that is performed simultaneously with the establishment of laser oscillation. By modulation, it changes the resonance parameters of the laser cavity to achieve changes in the output frequency of the laser, mainly including modulating the optical length of the resonant cavity or changing the gain loss spectrum position in the cavity;
external modulation technology refers to a technology that modulates the light field parameters using a modulator on the optical path of the laser exit after the laser oscillation is established. No matter which type of
is still in the exploration stage.
Most light sources with good tuning ability are not stable enough, and most stable light sources cannot be wide and tunable.
From the perspective of modulation method, the internal modulation method directly changes the resonant cavity parameters, and obtaining a large tuning range is relatively easy. However, due to the existence of laser establishment time, the instantaneous line width of the output frequency modulation light is relatively wide, resulting in a reduction in the coherence length of the light source. Or in order to establish a stable light field, the tuning rate must be limited.
external modulation method can quickly change the instantaneous frequency of the light field while maintaining the excellent characteristics of the seed light through tuning mechanisms such as electro-optical effect. However, due to the limited working bandwidth of the electro-optical effect itself, the increase in the tuning range of the light source is limited, that is, the highest resolution that can be achieved by the system.
Currently, the industry prefers external modulation method, which has the disadvantages of high cost and large size. The disadvantage of
FMCW is that it has high cost, and all its components need to have ultra-high accuracy, because the tuning frequency is at the THz level, which requires measuring instrument-level components.
is a component supplier, and each component requires high-precision detection and low yield. Even if
is mass-produced in the future, the cost will remain high. All optical surfaces must be within tighter tolerances, such as λ (wavelength) /20.
FMCW The ADC conversion rate requirement is 2 to 4 times that of ToF systems, and the accuracy requirements are higher.
requires FPGA to be able to receive data and perform ultra-high-speed FFT conversion.
Even with ASIC, the processing system complexity (and cost) required by the FMCW system is ten times that of the ToF system.
In addition to cost, FMCW does not have interference from external factors, but it will bring new interference itself.
is the same as millimeter wave radar. FMCW lidar needs to consider the interference of side lobe . FMCW system relies on side lobe suppression based on window function to solve the self-interference (clutter) . This interference is far less robust than the ToF system without side lobes.
To provide background information, a 10-microsecond FMCW pulse can propagate radially within a range of 1.5 kilometers.
In this range, any object will fall into the fast Fourier transform (time) sidelobe. Even the shorter 1 microsecond FMCW pulses may be destroyed by high-intensity clutter 150 meters away.
The side lobe of the first rectangular window Fast Fourier transform (FFT) is known to everyone -13dB, which is much higher than the level required to obtain a high-quality point cloud.
In addition, FMCW lidar has a slight delay problem, which is a natural defect of coherent detection and cannot be changed.
Aeva's main partners are Audi and ZF.
uses FMCW lidar. Other companies include Strobe, which General Motors acquired in 2017. This company has not taken any action since its acquisition.
Four Blackmore lidars are installed on the roof. The cost is as high as US$400,000
and the other is Blackmore, which was invested by BMW i Venture, which was acquired by Aurora in 2019.
(5) Approximate to Flash Ouster
Strictly speaking, Flash lidar refers to a single flash (laser pulse) imaging lidar.
borrows the term in the camera industry, also known as global shutter lidar.
The generalized Flash lidar refers to a focal plane array imaging lidar. It does not necessarily require a global shutter, but also a local shutter.
The typical representative of the global shutter lidar product is the ASC company acquired by Continental Automobile in 2016. Compared with scanning imaging lidar, Flash lidar has no moving parts and is an absolute solid-state lidar that can meet the highest level of automotive specifications.
Scanning imaging requires scanning the entire workplace to provide image (point cloud) , and the frame rate is usually 5-10Hz.
This means there is a delay of at least 100 milliseconds, which is unacceptable in high-speed scenarios.
If the scanning lidar wants to increase the frame rate, it must reduce the horizontal angle resolution, which is contrary to the two.
is very simple. The faster the scan, the lower the resolution.
But Flash does not. In theory, its pulse is only a few dozen nanoseconds to 1 nanosecond, which means that the frame rate can reach tens of KHz, or even 1MHz.
Of course, considering the data processing capabilities, the current Flash lidar is still 3 0Hz, but it can be said to be delay-free.
The HFL110 Flash lidar of mainland German cars has been confirmed to be used by Toyota L3 class unmanned mass-produced vehicles using
Although Toyota invested in Luminar, it still used the lidar of mainland German cars.
Flash The disadvantages of lidar are obvious: the power density of
is too low, which makes its effective distance generally difficult to exceed 50 meters, and the resolution is relatively low. Using high-power VCSEL and SPAD can solve some problems, but the cost has also increased rapidly.
Balanced German mainland cars are balanced between performance and cost, and their cost is estimated to be no more than 00 USD . After mass production, it can be reduced by about 100 USD.
In order to solve the disadvantages of signal-to-noise ratio and effective distance, some companies have made improvements to Flash lidar.
The improved design uses a VCSEL laser emission array, which is manufactured by semiconductor process chip technology. The current conduction of each small unit can be controlled, allowing the light emitting unit to conduct and light up in a certain mode, which can achieve the effect of the scanner and accurately control the scanning shape.
For example, when the vehicle speed is high, the FOV is reduced and the scanning accuracy is improved.
When the vehicle speed is low, the FOV will be increased and the detection range will be increased. Both
Ibeo and Ouster are designed like this.
Ibeo believes that this is a scanning lidar.
, and Ouster thinks it is a Flash lidar, but Multi-Segment was added before it.
In fact, both are the same lidar.
Ibeo has been working in the field of lidar for more than 20 years, and its Flash lidar performance is excellent. Except for the pixel count slightly lower than Lumianr, most of the remaining indicators are comparable.
But the reliability is far above Luminar, and the car rules are easier to pass.
. Why are these super giants optimistic about the Flash route?
I think the development direction of lidar is Flash, which can also be called depth camera .
says this because Flash lidar:
- is the easiest to use strict automotive regulations
- minimum volume
- installation location most flexible
- full chip
- lowest cost (the unit price can be easily achieved below US$100)
- performance mining potential (the depth camera is similar to the CMOS image sensor that just sprouted back then, and eventually replaced CCD)
Global technology community is in the global The R&D investment in the field is far higher than other types of lidars, all of which are super giants:
Broadcom (Tesla's partner) , Sony , Samsung, Apple , STMicroelectronics, Infineon, AMS, Lumentum, Toshiba , Panasonic, Canon , Hamamatsu, ON Semiconductor, Denso and Toyota are all developing Flash vehicle lidars.
In the field of optoelectronics:
Whether it is the SPIE International Society of Optoelectronics, OSA American Optical Society, ISSCC International Solid-State Circuits Association, European Optoelectronics Industry Association EPIC Conference, almost all papers are about the key components of Flash lidar, SPAD or VCSEL, and there is no traditional lidar paper at all.
depth cameras can be used not only in the automotive field, but also in other solid-state 3D sensing fields, as well as AR/VR.
Broadcom, the world's second largest IC design company that cooperates with Tesla to develop the next-generation chips, launched SPAD or SiPM array chips for in-vehicle Flash lidar at the EPIC online conference in November 2020. The upper and lower parts of the ultra-wide-angle lens of the iPhone 12 Pro form the lidar.
this with. There is no difference between the Flash LiDAR used in the car, and it is also a VCSEL+SPAD design, but the power is smaller and the size is smaller.
The mobile phone industry has actually widely used lidar, but it is called ToF camera.
Apple returns its real name.
Apple has already confirmed to build cars, so naturally it also needs to use its R&D achievements in the field of lidar, which can be used in the automotive field.
Apple iPad LiDAR disassembly, the sensor, that is, SPAD is provided by Sony .
Sony published a paper titled:
A 189×600 Back-Illuminated Stacked SPAD Direct Time-of-Flight Depth Sensor for Automotive LiDAR Systems at the ISSCC in December 2020, which also directly points to automotive Flash lidar.
Generally speaking, there are two elements that restrict the performance of Flash lidar:
- , one is the VCSEL emitted by laser. The other is the received SPAD
VCSEL. It is small in size, low in cost, easy to control, but relatively low in power.
Several VCSEL manufacturers are working hard to develop high-power VCSEL arrays. The fastest-progress is Lumentum, the main supplier of Apple , which is also the world's largest VCSEL manufacturer with a market share of about 45%.
Currently, the test products can achieve up to 10 watts of power, and 30 to 50 watts of power can be on par with non-Flash lidar. In terms of
automotive lidar SPAD, it currently only has 10,000 pixels, and 300,000 pixels in the mobile phone field are already the mainstream.
Japan has accumulated rich experience in the CCD field and has an overwhelming advantage in the SPAD field.
Canon has developed a 1-megapixel SPAD, which can easily crush the 128-line lidar with the highest performance, not to mention Luminar's MEMS lidar.
Samsung presented a paper titled:
A 4-tap 3.5μm 1.2Mpixel Indirect Time-of-Flight CMOS Image Sensor with Peak Current Mitigation and Multi-User Interference Cancellation at the International Solid State Circuit Seminar at the end of 2020, and proposed a 1.2 million pixel ToF sensor (i.e. SPAD) .
Panasonic has developed stacked SPAD. This SPAD can achieve an effective distance of 100 meters.
Toshiba is also developing chip-type SPAD.
Toshiba tried SPAD chip in March 2018, with a resolution of 240x96.
MEMS Lidar is just a transition product, but it is difficult to judge how long this transition period is.
I think it may take 3 years if it is fast, and it may take 6 years if it is slow. By then, the lidar will be installed in the rearview mirror like today's traditional cameras.
