Advantages and disadvantages of sensors and radar in driver assistance systems

    The European Microelectronics Summit was held in Paris not long ago, with the theme of "Prospects for V2V/V2I Technology Demand". Compared with the currently hotly debated autonomous vehicles, the focus of the conference was more on advanced driver assistance systems. According to the description of the summit host, both autonomous driving and basic V2V are inseparable from advanced driver assistance systems.

    STMicroelectronics has added a series of intelligent functions to its image sensor products. The company is currently developing a black and white image sensor that can sense red light, which will be used to detect traffic lights and taillights of the vehicles ahead.

    Freescale Semiconductor recently released a new product. The company said that the new product utilizes silicon germanium chip technology and is the most integrated system-level driver assistance system chip in the industry.

    Martin Duncan, director of STMicroelectronics' driver assistance systems and microcontroller division, believes that driver assistance systems will become an essential feature for both luxury and low-end cars.

    Euro NCAP and other organizations are also gradually including whether a new car is equipped with driver assistance systems as part of its scoring.

    However, all driver assistance systems rely on visual sensors or cameras. Based on this, the meeting also discussed the advantages and disadvantages of these two types of devices in driver assistance systems.

Radar vs. Sensor

    Automakers have a variety of options when developing driver assistance systems, including short-range radar, long-range radar, ultrasonic sensors and vision sensors.

    Freescale and STMicroelectronics chose different solutions, with the former focusing on radar equipment and the latter preferring to develop visual sensors. Euro NCAP has not made clear regulations on which type of configuration to use for driver assistance systems, whether to use sensors or radars.

    Car manufacturers are free to choose which configuration to use, and even combine radar and sensors - which is also the approach currently adopted by many car manufacturers.

    STMicroelectronics pointed out that radar has functional limitations in certain situations. For example, it cannot detect lane markings, traffic lights, pedestrians and control the vehicle's low/high beam switching.

    On the other hand, visual sensors also have their disadvantages. Even if they can still distinguish road conditions in the evening and on rainy days, they are not competent in foggy and snowy days. Dust and snowflakes can make visual sensors "blind". In addition, the detection range of sensors is not as wide as radar. Short-range radar can detect objects within 30 meters, while long-range radar has a further detection range of 150 meters.

    When it comes to the multifunctional requirements of driver assistance systems, STMicroelectronics believes that if a car must use a camera to monitor road conditions, then the advancement of visual sensor technology will be able to do what radar excels in. Its latest image sensor products can prove this.

STMicroelectronics monochrome image sensor

    STMicroelectronics has collaborated with Mobileye to develop a vision processor chip using Mobileye's advanced artificial vision technology. This chip is currently installed in the automotive products of BMW, GM, Volvo, Ford and other automakers.

    Another black-and-white image sensor is designed for automotive forward-looking cameras used in collision avoidance systems. This image sensor can distinguish red light, so it can detect traffic lights and the taillights and brake lights of vehicles ahead.

    In addition, when it acquires the video image directly in front of it, the image will not be compressed and deformed. However, when capturing a 360° panoramic image, it will be compressed. Regarding this, STMicroelectronics explained: "Currently, the images in cars are mainly compressed using Motion JPEG, but in the future, with the use of the new H264 compression standard, the image quality will be better."