Unmanned technology is now very mature now, to the current level of technology to see, if the complex urban traffic conditions into a laboratory-specific pattern, the scene has a unified system of vehicles and compliance with the normal pedestrian traffic, Then do not have the steering wheel, the whole car automatically driving the moment can be available. The problem lies Sensor in how the car can be on the reality of the complex traffic conditions well known, how can people like the eyes and brain as flexible. The key is to work with a variety of sensors to solve, they will eventually monitor the data passed to high-precision processors, identify roads, signs and pedestrians, to make acceleration, steering, braking and other decision-making.
In the intelligent perception part, the vehicle optical system and the car radar system are the most important to ensure the safety of traffic. At present, the mainstream sensor for the surrounding environment is LIDAR, millimeter wave, Sensor three kinds.
The following are the same as the '
By scanning the laser light reflected from an object to determine the distance of the object, you can form a 3D environment with a precision of up to centimeters, so it plays an important role in ADAS (advanced driving assistance system) and unmanned system. From the current car laser radar, the mechanical multi-beam laser radar Fuel Rail Pressure Sensor is the mainstream program, but subject to the high price factors have not yet universal.
In December 10 last year, the successful road test Baidu unmanned car body, in addition to the deployment of the millimeter-wave radar, video and other sensors, the roof of the placement of a larger, 70 million yuan worth of 64-bit laser Radar (VelodyneHDL64-E), Google is also using the same high-end configuration of the lidar. The advantages and disadvantages of the vehicle laser radar system mainly depends on the performance of 2D laser scanner. The more the laser transmitter harness is, the more cloud points are collected per second. However, the more the wiring harness represents the cost of the lidar is more expensive. Take Velodyne's product as an example, the price of a 64-wire harness lid is 10 times that of a 16-wire harness. Lidder in addition to high cost, encounter smoke fuel metering valve media and rain and snow in the performance of the general, will be the constraints of its play.
But as a core sensor, low-cost solution will accelerate the arrival of unmanned. At present, high precision automotive lidar products manufacturers are mainly concentrated in foreign countries, including the United States Velodyne, Quanegy and Germany's IBEO companies. Domestic laser radar products are relatively backward, China Aviation Automotive Systems Holdings Limited senior professional Zhou Shining has said that Bosch, the mainland, Valeo, Infineon, Delphi and other foreign parts enterprises have already seized the commanding heights of ADAS technology, especially in Sensor market layout, China's auto parts enterprises have lost at the starting line.
The following are the same as the '
Millimeter wave refers to the 30 ~ 300GHz frequency domain (wavelength 1 ~ 10mm) of electromagnetic waves, millimeter wave wavelength between the centimeter wave and light waves, so the millimeter wave both microwave guidance and photoelectric guidance advantages. Compared with the centimeter waveguide, the millimeter waveguide has the characteristics of small volume, light weight and high spatial resolution. Compared with the optical seeker such as infrared, laser and TV, the millimeter waveguide has the ability to penetrate the fog, smoke and dust, and has the characteristics Temperature Sensor of all-weather (except the rainy day), which can produce with the laser radar Complementary. In addition, the millimeter waveguide leader anti-interference, anti-stealth ability is better than other microwave seeker.
The disadvantage is that the millimeter-wave radar due to wavelength reasons, the detection range is very limited, can not perceive pedestrians, and laser radar can be all the surrounding obstacles for accurate modeling. In order to overcome the shortcomings of different shortcomings, car manufacturers are bound to these sensors together.
As ADAS function, often through the sensor + processor bundled way, the domestic automotive millimeter-wave radar system chip, the algorithm also mainly rely on imports, the cost is high. Accelerate the development of domestic 77GHZ millimeter-wave radar chip and as soon as possible vehicle applications, will be China's automotive millimeter wave radar industry opportunities.
The following are the same as the '
ADAS application camera as the main sensor is because the camera resolution is higher than other sensors, you can get enough environmental details to help the vehicle for environmental awareness, car camera can describe the appearance and shape of objects, read signs, etc. These features other The sensor can not do it. From the point of view of reducing costs, the camera is one of the powerful candidates for identifying sensors. Of course, the camera is the best choice in all clear cases, but it is influenced by environmental factors and external factors, such as lack of light in the tunnel, weather factors Resulting in narrowing the line of sight. An important tool for collecting image information, some similar road ABS Sensor signs recognition, lane line sensing and other intelligent functions from the camera to achieve. The current application of the camera are: monocular camera, rear view camera, stereo camera or binocular camera, look around the camera, according to 2015 more than 80 million new vehicles and bicycles 6-8 camera demand, the overall demand is expected in the future Over 600 million, corresponding to 100 billion market space.
Visual algorithms are indispensable in the ADAS technology route. The active sensor of the millimeter-wave radar has a low degree of dependence on the algorithm. The algorithm is relatively simple. The passive sensor of the camera has a high degree of reliance on the algorithm, which is usually provided by the third party. Such as visual algorithm business Mobileye.
Previously, Tesla and Mobileye ended the cooperation of the uproar, Mobileye provides a standard sensor installation + map data cloud service + software architecture platform to build, but Tesla is to collect data through the crowdsourcing mode to optimize the automatic driving experience, EyeQ3 chip limited Tesla to build their own maps, so Tesla future to develop their own image algorithms and image processing chip.
But Mobileye or with several products occupied 90% of the market share, this Israeli business in the local rivers and lakes status equivalent to BAT. Algorithms and hardware are the core of the ADAS system and are the core competencies of Mobileye. Mobileye stands at the top of ADAS, throwing rivals several streets, ADAS and traditional car vision products on software technology, hardware requirements are different, the traditional car electronics companies to cut into the ADAS market is not easy. It would take at least three or four years for a domestic team to complete the initial accumulation of technology from scratch.
Of course, in order to improve the accuracy of environmental perception, usually need a variety of sensor combinations, and ultimately provide a stable and durable solution. Now the typical Speed Sensor is the millimeter-wave radar, laser radar and car camera, other ultrasonic technology and infrared technology and the integration of these technologies will make the sensor industry to bring a huge market. However, no doubt, in the car to achieve full intelligence in the past few years, the sensor industry chain should be the first to get the benefits.