Advanced Driver Assistance Systems, in particular those to be used in future autonomous driving systems, rely on a variety of sensors - radar, lidar and cameras. Among them, radar typically takes the part of the long- and medium-distance object detection whereas lidar and cameras provide very detailed data of the surroundings, albeit at relatively low distance. The drawback of radar systems hitherto was that they return a signal that engineers refer to as "point cloud" - a signal that contains rather exact information with respect to distance and relative speed of the object detected but not many details as to its shape.
A phased-array radar system could significantly improve the image definition of the radar signal. Due to their high price, such systems today can be found only in areas where cost is not really a paramount issue - avionics, military technology, satellites etc. The San Diego research team, led by professor Gabriel Rebeiz, focused on integrating the functional elements of a phased-array radar system on a single chip. According to Rebeiz, the system markets the first time that RF beam forming phased-array technologies, basically coordinated groups of directional antennas, have been integrated into radar systems for civilian vehicles.
Developed in a coordinated effort by the UC San Diego Jacos School of Engineering, the Toyota Techncial Center, Fujitsu-Ten and the Michigan Technological Research Institute, the new Automotive Phased-Array Radar (APAR) is implemented as a single SiGe RF chip operating at 77 GHz, one of the frequency bands typically used for automotive radar systems. The system also has a build-in self-test feature on it (which by the way won the IEEE 2014 Microwave Prize).
The APAR delivers a high-resolution image of the vehicle surroundings at distances up to 100 metres. According to the university, the information is far more detailed than what is available from today's automotive radar systems. It can be used in utilised in lane changing assistants as well