Will LiDAR trigger the car's rain sensor to turn on?

Will LiDAR trigger the car's rain sensor to turn on? This article will reveal the secret. A San Francisco netizen recently said online, "I live next to a parking lot used for autonomous driving testing , and I recently realized that when I get close to a car using LiDAR , it triggers my rain sensor and the wipers start running. It has happened several times before, and I only recently connected it to LiDAR. This seems to be an unexpected bug. Does anyone know about this?"

Everyone who studies autonomous driving is familiar with LiDAR. The basic principle is that the pulsed laser emitted by the laser is incident from the air to the ground, hitting trees, roads, and cars, causing scattering. Part of the light wave will be reflected back to the LiDAR receiver, and the distance to the target is calculated based on the time difference of the pulse signal recorded by the timer

Can it affect the rain sensor? Let's take a look at the principle of the rain sensor. The rain sensor is mainly used to detect whether it is raining and the amount of rain. When the car is driving in bad weather such as rain and snow, the rain sensor reports to the domain controller, and the controller drives the wipers, lights, and windows to complete the early warning action for rainy days. In theory, this is an ADAS function. To ensure that the driver has a good view on rainy days, on rainy days, when the sensor detects that rain has fallen on the windshield, it will issue a command to the wipers to start working, and the rain on the front windshield of the car will be cleared. The intelligent control system detects and analyzes the frequency of the wiper to improve the driver's visual effect, ensure unobstructed vision, and ensure driving safety.

Common rain sensors include flow type rain sensors, electrostatic rain sensors, piezoelectric rain sensors, and infrared raindrop sensors. Of course, there are also special ones, such as Tesla using image vision to judge whether it is raining. Among them, flow type, electrostatic type, and piezoelectric rain sensors are strongly related to the physical properties of water droplets, and are less likely to be affected by laser pulses. According to the description of netizens, if there is an impact, it is speculated that it may be an infrared raindrop sensor.

This sensor relies on the light emitted by the light emitting diode to be adjusted by the lens system and then irradiated onto the windshield in a parallel light state; when the glass is dry, the light will be totally reflected and received by the receiving device in a parallel light state through the lens system, with an output maximum value of 100%; when there is rain or raindrops on the glass, the light will not be totally reflected due to the change in refractive index, but will be partially reflected depending on the size of the water droplet area. At this time, the receiving tube only receives part of the signal, and the amount of rainfall can be calculated according to the percentage ratio.

The sensor is affected by background light, the photoelectric conversion signal is weak, and the signal is easily submerged by noise. When it comes into contact with a working lidar, there is a certain probability that signal noise will be generated to the sensor, causing it to produce an erroneous response.

The sensor based on light intensity change is installed on the inside of the car windshield, without direct contact with rain. Under the protection of the windshield, it can work stably for a long time, so infrared scattering rain sensors are becoming mainstream products. After all, vehicles with laser radar are not yet popular, so the incident of this netizen rarely happens in daily life. It can be seen that in the process of ADAS transitioning to autonomous driving, the integration of multi-module functions must be considered more carefully to avoid unnecessary "conflicts".

When it comes to rain sensors, we have to mention Tesla's solution of visually identifying raindrops. Tesla Autopilot can identify wet weather through image recognition technology, and then turn on the wipers when necessary. This solution seems smarter, but it is actually unreliable. According to research by Tencent Keen Lab, by cleverly generating an image in the physical world, the system will be disturbed and return an "incorrect" result, and then awkwardly turn on the wipers.