Follow
Subscribe

All-range, Full-stack 4D Radar Could Improve Safety in Driverless Cars

Home > News > Content

【Summary】To clarify, radar, which stands for radio detection and ranging, is not obsolete and cannot be replaced by LIDAR. Radar is specifically applicable at close ranges, for parking or lane-changing protocols.

Michael Cheng    Apr 06, 2017 5:55 AM PT
All-range, Full-stack 4D Radar Could Improve Safety in Driverless Cars

When it comes to sensing technology used on self-driving cars, all the hype and attention has been around LIDAR. But driverless vehicles use a handful of other components, like cameras and radar. To clarify, radar, which stands for radio detection and ranging, is not obsolete and cannot be replaced by LIDAR. The two types of sensors have very different applications. 

To compare the two technologies, LIDAR is useful for 3D mapping at distances beyond 100 meters. It relies on laser light to generate points in the environment, while consuming heaps of data. Radar is specifically applicable at close ranges, for parking or lane-changing protocols. This type of sensor consumes less data but offers less angular accuracy than LIDAR.

When used together, both units cross reference each other to validate motion predictions and image generations on busy roads.

 mobileye-delphi-sdc-hero.jpg

Radar and SAE-L4

Now that we've established the importance of radar systems in autonomous driving, let's take a closer look at its functionality and benefits. Radar will be pushed to its limits, as the component takes on a more critical role in SAE-L4 driverless platforms. With less human intervention and increased reliance on image processing, radar must be able to cope with a wide range of weather conditions, which it can do better than LIDAR (at close distances).

"To fit the radar into the flat bumper assembly and its limited space, it is necessary to use highly integrated design, including using part of the radar subsystem PC board as its antenna," explained Bill Schweber from Mouser Electronics.

Radars located at the front, side and rear bumper of vehicles regularly operate at a frequency of 77GHz. From this angle, the unit can "see" nearby vehicles and people approaching the car. This becomes incredibly useful when performing hairline parking maneuvers. Going back to bad weather, some individuals have reported radar sensors to work in extreme snowy conditions.

ADAS features, including forward collision sensing, blind-spot detection and adaptive cruise control, does not seem to be hindered by rough weather – confirmed by Sam Abuelsamid, a senior analyst for Navigant Research, a marketing intelligence firm based in Boulder, Colorado.

 maxresdefault.jpg

Next-generation 4D Radar Systems

Next-generation radar technology will make driverless cars exponentially safer, when around other vehicles. Arbe Robotics, with the release of Ultres, an all-range 4D radar being developed by the startup, is leading the way in radar testing. The group's solution is capable of detection up to 300 meters with a refresh rate of 50 times per hour. With its 3D imaging capabilities, it could one day replace cameras. But in order to do that, radar must be able to "see" color, which it currently cannot. Cameras are the cheapest self-driving sensor out of the three and is mostly used for scene interpretation, image classification and texture sensing.

"Radar is a proven technology increasingly becoming more efficient for the autonomous car. The new RFCMOS technology recently introduced to the market will allow smaller, lower power, efficient sensors that fit right into the OEM cost reduction strategy," said Davide Santo is a 17-year veteran of Motorola, Freescale ad NXP, ADAS microcontroller and processor Product Line General Manager.

Prev                  Next
Writer's other posts
Comments:
    Related Content