Car Sensors - A detailed explanation of LiDAR

Introduction: LiDAR is mainly used in autonomous driving applications to detect obstacle information on the road, transmit data and signals to the brain of autonomous driving, and then make corresponding driving actions. However, common outdoor interference factors such as rain, fog, snow, dust, high and low temperatures have a great impact on the recognition of LiDAR. Therefore, before LiDAR is put into practical application, a large number of tests need to be carried out in special environments such as rain, fog, light, and dust. Under natural environmental conditions, the required test scenes are hard to come by and cannot be reproduced, and cannot meet the needs of a large number of real-world tests of LiDAR. On this basis, the National Intelligent Connected Vehicle (Changsha) Test Area is about to complete typical test scenes such as rain, fog, light, and dust. It is also the only test area in China that has scenes such as rain, fog, light, and dust at the same time, which can meet the needs of large-scale reproducible real-world tests of sensors such as LiDAR. (The following content is reproduced from Today's Radio and Television, please delete if infringed)

Figure 1 LiDAR is affected by the environment

LiDAR is a sensor that is currently changing the world. It is widely used in self-driving cars, drones, autonomous robots, satellites, rockets, etc. The laser measures the propagation distance (Time of Flight, TOF) between the sensor transmitter and the target object (as shown in Figure 2), analyzes the reflected energy size, amplitude, frequency and phase of the reflected spectrum on the surface of the target object, and outputs a point cloud, thereby presenting accurate three-dimensional structural information of the target object.

Figure 2 LiDAR ranging and point cloud

LiDAR is composed of a laser transmitting unit and a laser receiving unit. The transmitting unit works by emitting layers of laser beams. The more layers there are, the higher the accuracy (as shown in Figure 3), but this also means that the sensor size is larger. After the transmitting unit emits the laser, it will be reflected when it encounters an obstacle and then be received by the receiver. The receiver creates a set of point clouds based on the time when each laser beam is emitted and returned. A high-quality LiDAR can emit more than 200 laser beams per second.

Figure 3 Laser point clouds formed by different laser beams

As for the wavelength of laser, currently, laser transmitters with wavelengths of 905nm and 1550nm are mainly used. Light with a wavelength of 1550nm is not easy to transmit in the liquid of the human eye. Therefore, 1550nm can greatly increase the transmission power while ensuring safety. High power can achieve a longer detection distance, and long wavelength can also improve anti-interference ability. However, 1550nm lasers require the use of InGaAs, which is currently difficult to mass produce. Therefore, Si materials are currently used more to mass produce 905nm LiDAR. Safety is ensured by limiting power and pulse time.

1. The structure of LiDAR

The key components of LiDAR include control hardware DSP (digital signal processor), laser driver, laser emitting diode, transmitting optical lens, receiving optical lens, APD (avalanche optical diode), TIA (variable transconductance amplifier) ​​and detector according to the signal chain of signal processing, as shown in Figure 4. Except for the transmitting and receiving optical lenses, all other components are electronic components. With the rapid evolution of semiconductor technology, the performance is gradually improved while the cost is rapidly reduced. However, optical components and rotating machinery account for most of the cost of LiDAR.

Figure 4 Key components of LiDAR

2. Types of LiDAR

There are different types of lidar on the market, which can be divided into mechanical, MEMS, phased array, and floodlight array (FLASH) according to the driving method.

Mechanical

Take the 64-line radar launched by Velodyne in 2007 as an example. It stacks 64 lasers vertically together and rotates at a speed of 20rpm. Simply put, the laser points are turned into lines through rotation, and the lines are converted into surfaces through 64-line stacking to obtain point cloud data to obtain 3D environmental information.

The mechanical structure requires a complex mechanical structure, and the point cloud measurement requires precise positioning of the installation. Considering the impact of the environment and aging, the average failure time is only 1000-3000 hours, which is difficult to meet the minimum requirement of 13,000 hours by the car factory. And because the LiDAR is installed on the roof, the civilian field needs to consider external maintenance issues, such as the impact of car washing. Therefore, the mechanical structure greatly limits the cost and application promotion.

MEMS

MEMS lidar uses micro-electromechanical system technology to drive a rotating mirror to reflect laser beams in different directions.

The advantages of solid-state LiDAR include: fast data acquisition speed, high resolution, strong adaptability to temperature and vibration; through beam control, the detection points (point cloud) can be arbitrarily distributed, for example, on highways, the front is mainly scanned, the side is sparsely scanned but not completely ignored, and the side scanning is strengthened at intersections. Mechanical LiDAR, which can only rotate at a constant speed, cannot perform such delicate operations.

Typical applications include Valeo SCALA LiDAR, which is currently used in Audi A8 (the first L3 autonomous driving vehicle). Installed in the front bumper, it uses MEMS technology to achieve a scanning angle of 145° and a detection distance of 80m.

Figure 5 Audi A8's LiDAR

Phased Array (OPA)

Principle of optical phased array radar: It mainly uses the principle of light interference. The position of the central bright fringe (main lobe) after grating diffraction can be changed by changing the phase difference of the incident light in different slits, as shown in the figure below.

Figure 6 Principle of phased array radar

Advantages and Disadvantages of Phased Array (OPA)

advantage:

①Simple structure and small size: Since no rotating parts are required, the structure and size of the radar can be greatly compressed, the service life can be increased, and the cost can be reduced.

② Simple calibration: Mechanical LiDAR has a fixed optical structure, so its position and angle often need to be precisely adjusted to suit different vehicles. Solid-state LiDAR can be adjusted through software, greatly reducing the difficulty of calibration.

③ Fast scanning speed: Not restricted by the speed and accuracy of mechanical rotation, the scanning speed of the optical phased array depends on the electronic properties of the materials used, and can generally reach the MHz level.

④ High scanning accuracy: The scanning accuracy of the optical phased array depends on the accuracy of the control electrical signal, which can reach more than one thousandth of a degree.

⑤Good controllability: The beam pointing of the optical phased array is completely controlled by electrical signals. It can be pointed arbitrarily within the allowed angle range and can perform high-density scanning in key areas.

⑥Multi-target monitoring: A phased array surface can be divided into multiple small modules, and each module can be controlled separately to lock and monitor multiple targets at the same time.

shortcoming:

① Limited scanning angle: Adjusting the phase can only change the central bright pattern by about ±60° at most. To actually achieve 360° acquisition, generally 6 phases are required.

② Sidelobe problem: In addition to the central bright fringe, grating diffraction will also form other bright fringe. This problem will cause the laser to form side lobes outside the maximum power direction, dispersing the laser energy.

③ High processing difficulty: Optical phased array requires that the size of the array unit must not be greater than half a wavelength. Generally, the working wavelength of current laser radars is about 1 micron, so the size of the array unit must not be greater than 500nm. Moreover, the higher the array density, the more concentrated the energy, which increases the requirements for processing accuracy and requires certain technological breakthroughs.

④ Large receiving surface and poor signal-to-noise ratio: Traditional mechanical radar only requires a very small receiving window, but solid-state laser radar requires an entire receiving surface, which will introduce more ambient light noise and increase the difficulty of scanning and analysis.

FLASH

The principle of the flood array is similar to that of a TOF camera, that is, a flash. Unlike MEMS or OPA solutions, it does not scan, but directly emits a large area of ​​laser covering the detection area in a short time, and then uses a highly sensitive receiver to complete the image of the surrounding environment. It operates more like a camera. The laser beam will diffuse directly in all directions, so only one flash can illuminate the entire scene. The system will then use a micro-sensor array to collect laser beams reflected from different directions.

Figure 7 Floodlight LiDAR

One of the major advantages of Flash LiDAR is that it can quickly record the entire scene, avoiding various troubles caused by the movement of the target or LiDAR during the scanning process. There are two current development directions: one is the single-photon counting type of Geiger mode APD to directly generate digital images by counting photons; the other is the traditional CMOS light intensity simulation acquisition to obtain an intensity map and convert the intensity map into distance information.

3. Data transmission of LiDAR