# Lidar: Automatic Driving from the Perspective of Photovoltaic Technology
Lidar and its competing sensor technologies (cameras, radar, and ultrasound) have strengthened the need for sensor fusion, and have also placed higher demands on careful selection of photodetectors, light sources, and MEMS galvanometers.
Advances in sensor technology, imaging, radar, light detection technology and ranging technology (laser radar), electronic technology, and artificial intelligence have enabled dozens of advanced driver assistance systems (ADAS) to be realized, including collision avoidance, blind spot monitoring, Lane departure warning and parking assistance. The synchronized operation of these systems through sensor fusion allows a completely autonomous vehicle to monitor the surroundings and warn the driver of potential road hazards, even taking driver-independent evasive action to prevent collisions.
Self-driving cars must also distinguish and identify objects in front of them at high speeds. With ranging technology, these self-driving cars must quickly build a three-dimensional (3D) map of approximately 100m distance and create high-angle resolution images up to 250m. If the driver is not present, the vehicle's artificial intelligence must make the best decision.
One of the basic methods for accomplishing this task is to measure the round-trip flight time (ToF) of the energy pulse from the autonomous vehicle to the target and back to the vehicle. When you know the speed of the pulse through the air, you can calculate the distance to the reflection point - the pulse can be ultrasound (sonar), radio wave (radar) or light (lidar).
Of these three ToF technologies, LiDAR is the best choice for providing higher-angle resolution images because it has smaller diffraction characteristics and beam divergence and can better identify adjacent objects than microwave radars. This high-angle resolution is particularly important at high speeds, providing enough time to deal with potential hazards such as head-on collisions.
#Selection of laser source
In the ToF Lidar, the laser emits a pulse of duration τ, which triggers the internal clock in the timing circuit (shown below) at the moment of launch. The light pulse reflected from the target reaches the photodetector, and the conversion generates an electrical signal output to stop the clock. This way of measuring the time to and from ToF Δt can calculate the distance R to the reflection point.
If the laser and photodetectors are actually in the same location, the distance is determined by the following formula:
Medium c is the speed of light in vacuum, n is the refractive index of the propagating medium (about 1 for air), and there are two factors affecting the distance resolution ΔR: the uncertainty δΔt at measuring Δt and the space caused by the pulse width Error w (w = cτ).
The first factor represents the ranging resolution ΔR=1/2 cδΔτ, while the second represents the ranging resolution ΔR=1/2 w = 1/2 cτ. If the distance is measured with a resolution of 5 cm, the above relation means that δΔt is about 300 ps and τ is about 300 ps, ​​respectively.
Time-of-flight laser radar requires photodetectors and subsequent electronics to have little time jitter (a major contributor to δΔτ) and pulsed lasers that can emit short pulse widths, such as relatively expensive picosecond lasers. The current laser in a typical automotive lidar system produces a pulse of about 4ns duration, so reducing the beam divergence is necessary.
The beam divergence depends on the ratio of the wavelength and the size of the transmitting antenna (microwave radar) or the lens aperture size (laser radar). Microwave radar has a large ratio, so it has a larger divergence and a lower angular resolution. The microwave radar (black) in the figure will not be able to distinguish between the two cars, but the laser radar (red) can.
One of the most critical choices for automotive laser radar system designers is the wavelength of light. There are several factors that limit this choice:
· Safety to human vision
· Propagation characteristics in the atmosphere
Laser availability and availability of photodetectors
The two most popular wavelengths are 905 and 1550 nm, and the major advantage of 905 nm is that silicon absorbs photons at this wavelength, while silicon-based photodetectors are generally much better than indium gallium arsenide (InGaAs) near-infrared detectors needed to detect 1550 nm light. Cheaper.
Hamamatsu near-infrared MPPC (silicon photomultiplier tube) , which can be used for automatic driving laser radar, has high detection efficiency at 905 nm, fast response, wide operating temperature range, and is suitable for laser radar applications in various situations, especially Long-distance measurement using TOF ranging method.
Click for details: Lidar technology in driverless driving - Hamamatsu's Laser Eye
However, the 1550nm human vision is much safer, and it is possible to use a laser with a single pulse of higher radiant energy—an important factor in the choice of wavelength of light.
1550nm Detector Hamamatsu InGaAs APD G8931
Atmospheric attenuation (under all weather conditions), scattering of particles in the air, and the reflectance of the target surface all depend on the wavelength. Due to the wide variety of possible weather conditions and reflective surfaces, the choice of wavelength for automotive laser radar under these conditions is a complex issue. In most practical cases, the light loss at 905 nm is smaller because the moisture absorption at 1550 nm is greater than at 905 nm.
#Selection of photodetectors
Only a fraction of the pulsed photons can reach the active area of ​​the photodetector. If the atmospheric attenuation does not change along the pulse path, the divergence of the laser beam is negligible, the spot size is smaller than the target, the incident angle is normal to the detector and the reflector is Lambertian (reflects in all directions), then the light receiving peak power P(R) )for:
P0 is the peak power of the emitted laser pulse, Ï is the reflectivity of the target, A0 is the aperture area of ​​the receiver, η0 is the transmittance of the optical system, and γ is the atmospheric extinction coefficient.
The equation shows that as the distance R increases, the received power decreases rapidly. For a reasonable choice of parameters, R = 100 m, the number of returned photons in the active area of ​​the detector is approximately a few hundred to several thousand, and the photons emitted usually exceed 1012. These echo photons are detected at the same time as the background photons, and the background photons do not have any useful information.
Using a narrowband filter can reduce the background light reaching the detector, but it cannot be reduced to zero. The effect of background light reduces the detection dynamic range and increases the noise (background photon shot noise). It is worth noting that under typical conditions ground solar irradiance is less than 905 nm at 1550 nm.
The basic principle of the time of flight (ToF) lidar
Creating a complete 3D map in a 360° x 20° area around a car requires a beam splitter to be scanned after beam splitting, scanning with multiple laser beams, or covering the entire desired area and collecting it back Point cloud data. The former is referred to as scanning (Scanning) laser radar, which is referred to as flash (flash) lidar.
Using a narrowband filter can reduce the background light reaching the detector, but it cannot be reduced to zero. The effect of background light reduces the detection dynamic range and increases the noise (background photon shot noise). It is worth noting that under typical conditions ground solar irradiance is less than 905 nm at 1550 nm.
There are several ways to scan a laser radar . In the first method, using Velodyne as an example (San Jose, CA), a laser radar platform is installed on the top. The radar rotates at a speed of 300 to 900 rpm and emits 64 pulses of 905 nm laser light. Each beam has a corresponding avalanche photodiode (APD) detector. A similar approach is to use a rotating polygon mirror, each with slightly different angles of inclination, to guide reflection of a single pulsed beam at different azimuths and angles. The mechanical moving parts of both designs have the risk of failure when the external driving environment is harsh.
Hamamatsu New 100m Automatic Driving Lidar Detector 16ch Silicon APD S14137-01CR
A second, more compact, scanning laser radar uses a micro-electromechanical system (MEMS) galvanometer to electrically direct one or more beams in a two-dimensional direction. Although there are still moving parts (oscillating mirrors) in technology, the amplitude of the oscillation is small and the frequency is also high enough to prevent the mechanical resonance between the MEMS galvanometer and the car. However, the geometry of the galvo mirror limits its oscillation amplitude, which makes the viewing angle limited - this is a disadvantage of the MEMS method. However, this method has attracted people's attention due to its low cost and high availability.
Hamamatsu's Newest MEMS Mirror Products Just Showed at Shanghai Light Fair in Munich
Optical phased array (OPA) technology is the third laser radar technology to compete, and it is increasingly popular with reliable "fixed parts" designs. It consists of an array of optical antennas illuminated by coherent light. Beam steering is achieved by independently controlling the phase and amplitude of each unit's light emission, thereby interfering with the far field to produce the ideal illumination direction, from single beam to multiple beam variations. Unfortunately, the loss of light limits the usable range of various OPA components.
The flash lidar fills the target scene with light, and the illuminated area matches the detector's field of view. The detector is an array of APDs on the detection optical focal plane. Each APD independently measures the ToF of the image target feature on it. This is a true "no moving part" approach where the tangential (vertical, horizontal) resolution is limited by the size of the 2D detector pixel.
However, the main disadvantage of the flash lidar is the number of echo photons: once the distance exceeds tens of meters, the amount of returned light is too small to be reliably detected. If you do not cover all detection environments directly with light but use structured light (for example, a lattice format) and sacrifice a certain tangent resolution, you can increase the echo light intensity. In addition, vertical cavity surface emitting lasers (VCSELs) make it possible to emit several thousand beams of light simultaneously in different directions. Â
# Get rid of the limitations of the ToF method
ToF laser radar is vulnerable to noise due to its weak echo pulse and broad band width of detection electronics, and threshold triggering will produce measurement error of Δt. Therefore, a frequency-modulated continuous wave (FMCW) lidar is a very meaningful alternative.
In an FMCW radar or å•å•¾ modulation radar, the antenna continuously transmits radio waves of a frequency modulated. For example, as time T increases linearly from Æ’0 to Æ’max, then linearly decreases from Æ’max to Æ’0 as T. If the wave is reflected back to the emission point on a moving object within a certain distance, its instantaneous frequency will be different from the instantaneously transmitted radio wave. This difference is caused by two factors: the distance to the object and its relative radial velocity. The frequency difference can be obtained by electronic measurement and the distance and speed of the object can be calculated at the same time (see the figure below).
In radars, by measuring fB1 and fB2 electronically, the distance to the reflection target and its radial velocity can be determined.
Inspired by Lei Radar, FMCW Lidar can be obtained in different ways. In the simplest design, people can adjust light intensity brightly. This frequency is affected by the same regularity (such as the Doppler effect) of the carrier frequency of the FMCW radar. The returned light is detected by the photodetector and the modulation frequency is restored. The output is amplified and mixed with its own oscillation frequency to allow frequency shift measurement. And thus calculate the distance and speed of the target.
However, FMCW laser radar has certain limitations. Compared with ToF laser radar, it needs more computing power. Therefore, the speed is slow when generating a full three-dimensional surround image, and the measurement accuracy is linear to the time of modulation. Very sensitive.
Although it is challenging to design a fully functional lidar system, these challenges are not insurmountable. As research continues, we are getting closer to the era when most cars are fully automated after production.
references
1.J. Wojtanowski et al., Opto-Electron. Rev., 22, 3, 183-190 (2014).
Author / Hamamatsu Research Scientist: Slawomir Piatek Hamamson Marketing Engineer: Jake Li
The original link: LIDAR: A photonics guide to the autonomous vehicle market
Insulated Terminals,Terminals,High-quality insulated terminals
Taixing Longyi Terminals Co.,Ltd. , https://www.longyicopperterminals.com