LiDAR, also known as optical radar (Light Detection And Ranging) or laser radar is the abbreviation of laser detection and ranging system, which presents precise two-three dimensional structural information of the target by determining the propagation distance between the sensor transmitter and the target object and analyzing the reflected energy magnitude, amplitude, frequency and phase of the reflected wave spectrum on the surface of the target object.
LiDAR has been developed on a large scale since the 1960s, shortly after the laser was invented. Currently LiDAR manufacturers mainly use laser emitters with wavelengths of 905nm and 1550nm. Light with a wavelength of 1550nm is not easily transmitted in the fluid of the human eye, which means that LiDAR with a wavelength of 1550nm laser can be quite high in power without causing retinal damage. Higher power means longer detection distance and longer wavelength means easier penetration of dust haze. However, due to cost reasons, the production of LiDAR with a wavelength of 1550 nm requires the use of expensive gallium arsenide materials. Manufacturers prefer to use silicon materials to manufacture LiDARs close to the visible wavelength of 905nm and strictly limit the power of the emitter to avoid permanent damage to the eyes.
According to the different ranging methods, LiDAR is mainly divided into ToF (Time of Flight) and FMCW (Frequency Modulated Continuous Wave) types, and ToF type is the vast majority of LIDARs currently in mass production.
In the ToF (Time of Flight) method, the LiDAR transmitter emits a pulse, hits an object and returns, and the receiver receives the return wave and calculates the difference in reception time between the two, and multiplies it by the speed of light to achieve distance measurement between objects.
Figure: LiDAR method ToF (Time of Flight)
Coherent methods are also used, i.e., for frequency modulated continuous wave (FMCW) LiDAR transmitting a continuous beam with a frequency that varies steadily over time. Since the frequency of the source beam is constantly changing, differences in beam transmission distance result in differences in frequency. After mixing the echo signal with the local oscillation signal and low-pass filtering, the resulting differential frequency signal is a function of the beam round-trip time. The FM continuous wave LiDAR will not be interfered by other LiDAR or sunlight and has no range blindness; it can also measure the speed and distance of objects using Doppler shift. The FM continuous wave LiDAR concept is not new, but faces a number of technical challenges, such as the linewidth limitation of the emitted laser, the frequency range of the linear FM pulses, the linearity of the linear pulse frequency variation, and the reproducibility of individual linear FM pulses.
Figure: LiDAR method FMCW (Frequency Modulated Continuous Wave)
Amplitude-modulated continuous wave (AMCW) LiDAR is similar to the basic time-of-flight system in that the AMCW LiDAR emits a signal that measures the time of the laser reflection back. The difference, however, is that time-of-flight systems emit only one pulse, and AMCW LiDAR modulates by varying the polar current in the laser diode to adjust the intensity of the emitted light.
LiDAR mainly consists of laser transmitter, receiver, scanner, lens antenna and signal processing circuit. There are two main types of laser transmitters, one is laser diode, which usually has two substrate materials, silicon and gallium arsenide, and the other is the very hot vertical cavity surface emission (VCSEL) (such as LiDAR on iPhone), the advantages of VCSEL are low price, extremely small size and low power consumption, the disadvantage is that the effective distance is relatively short and requires multi-stage amplification to reach the effective distance for vehicle use.
LIDAR mainly applies the principle of laser distance measurement, and how to make a suitable structure so that the sensor can emit laser beams in multiple directions and how to measure the time of laser round trip distinguishes the different LIDAR structures.
As an example, Velodyne introduced a LiDAR in 2007 that stacks 64 lasers vertically so that the entire unit rotates many times per second. There is a physical sense of rotation in the transmitting and receiving systems, which means that the laser points are turned into lines by continuously rotating the transmitters and lining up multiple laser emitters in the vertical direction to form a surface for the purpose of 3D scanning and receiving information. However, due to the complex mechanical structure to achieve high-frequency accurate rotation, the average failure time is only 1000-3000 hours, which is difficult to meet the minimum 13000 hours requirement of vehicle manufacturers.
Figure: Velodyne LIDAR Structure
Some of the common mechanical radar products include those produced by German company Sick, Japanese company Hokuyo, and some companies make more affordable versions such as RPLiDAR.
The technology of micro-electro-mechanical system is used to drive the rotating mirror and reflect the laser beam to point in different directions. The advantages of hybrid solid-state LiDAR include: fast data acquisition, high resolution, adaptability to temperature and vibration; through beam control, detection points (point clouds) can be distributed arbitrarily, for example, on highways mainly scanning far ahead, sparse scanning for the side but not completely ignored, and enhanced side scanning at intersections. Mechanical LIDAR, which can only rotate at a uniform speed, is unable to perform such fine operations.
Figure: Solid-state LIDAR structure
Solid-state lidars can be smaller, lighter, and more durable than traditional lidar systems. And thus, are quite popularly applied in robot vacuum cleaners. Solid-state lidar technology is rapidly evolving, and new companies are emerging in this space such as the YDLIDAR.
Phased array transmitter consists of an array of several transmitting and receiving units. By changing the voltage loaded on different units and thus changing the characteristics of the light waves emitted from different units, independent control of the light waves from each unit is realized, and by adjusting the phase relationship between the light waves radiated from each phased unit, a mutually reinforcing interference is generated in the set direction to achieve a high intensity beam, while the light waves emitted from each unit in other directions The light waves from each unit in other directions cancel each other. The phased control units of the phased array can make one or more high-intensity beams scan in the airspace according to the design direction under the control of the program. More detail about phased array can be found at https://en.wikipedia.org/wiki/Phased_array
Figure: Phrase Array Principle of LIDAR
(sourced from: ALFA-Pi: Generic LiDAR Ethernet Interface)
But the optical phased array manufacturing process is more difficult, this is due to the requirement that the size of the array unit must not be larger than half a wavelength, the general current LIDAR task wavelength are in about 1 micron, which means that the size of the array unit must not be larger than 500 nanometers. And the more arrays, the smaller the size of the array unit, the more energy is concentrated to the main flap, which requires higher processing accuracy. In addition, the material selection is also a very critical element.
OPA lidar systems have the advantage of being able to generate high-resolution 3D images of the environment with no mechanical scanning components, which can make them more reliable and robust than traditional scanning lidar systems.
There are several companies that are developing or producing Optical Phased Array (OPA) lidar systems. Some examples include:
- Aeva: Aeva is a company that produces 4D lidar sensors for use in autonomous vehicles and other applications. Their lidar sensors use an OPA approach to generate high-resolution, long-range 3D images of the environment.
- Luminar: Luminar is another company that produces lidar sensors for use in autonomous vehicles. Their "Iris" sensor uses an OPA approach to generate a 3D point cloud of the environment with high resolution and long range.
- Baraja: Baraja is a company that produces lidar sensors for use in autonomous vehicles, smart cities, and other applications. Their "Spectrum-Scan" lidar system uses an OPA approach to generate high-resolution 3D images of the environment with no moving parts.
The principle of FLASH is that it does not scan, unlike MEMS or OPA solutions, but directly emits a large area of laser light to cover the detection area for a short time, an array of laser beams that are emitted simultaneously over a wide area to provide a 3D view of the surrounding environment. And then uses a highly sensitive receiver to complete the image of the surrounding environment. This approach is also known as "flash lidar" or "laser flash imaging."
In contrast to traditional scanning lidar systems, which emit a single laser beam and scan it across a scene to build up a 3D image, a flooded light array system can capture an entire scene in a single flash. The system measures the time it takes for the laser pulses to bounce off of objects in the environment and return to the sensor, which allows it to calculate the distances and locations of those objects in 3D space.
Flooded light array systems have the advantage of being able to capture a large amount of data in a short amount of time, making them useful for applications such as autonomous vehicles, robotics, and environmental monitoring.
One example is Quanergy, a company that produces 3D LiDAR sensors for use in autonomous vehicles, drones, and other applications. Their "M8-Prime" sensor is an example of a flooded light array system that uses a dense array of laser beams to capture a 360-degree view of the surrounding environment in real-time.
Another example is Ouster, which produces high-resolution 3D lidar sensors for applications in robotics, automation, and mapping. Their "OS1" sensor uses a similar flooded light array approach to capture a wide-angle view of the environment.
It's worth noting that these systems can be quite expensive and are typically only used in specialized applications. However, as the technology advances and becomes more widespread, we may see more affordable versions of these systems become available.