During the winter months, the cold weather brings new challenges for autonomous vehicles, especially for optical sensors since they lose effectiveness similar to the human eye.
Both the human eye and optical-based sensors use light as a primary source of input, which means they both fall to the same shortcomings when in the same situations. Radar-based sensors don’t have these problems, as they use radio wave-based input to collect data; this allows them to function better in situations where visibility is low and provide a much safer solution when faced with difficult weather conditions.
The Benefits of Using Radar-Based Sensors
With the myriad of possible weather conditions like snow, sleet, fog, etc., that autonomous vehicles need to be able to navigate through, it is important to ensure that the sensors embedded in the vehicle can provide a safe driving experience. Different weather conditions can reduce the visibility of more traditional optical sensors because airborne particles create unwanted camera artifacts that can be detected as targets. Additionally, glare from snow can effectively blind an optical-based sensor.
Radar-based sensors use much longer wavelengths that avoid these unwanted obstructions, allowing the sensor to continue to properly function even during intense weather. On top of that, radar-based sensors are much better at handling large blockages, such as mud, while optical sensors can malfunction or fail when presented with those obstructions. Radar-based sensors consistently provide a much safer driving experience while in suboptimal weather conditions.
Taking It Another Step
While providing more stable results than optical sensors, standard automotive radar sensors aren’t perfect and can still be triggered by false alarms. 4D Imaging Radar can solve this problem by using ultra-high resolution which is able to sort relevant, irrelevant and false targets. Unlike standard automotive sensors which measure objects in three dimensions, range, azimuth and doppler, 4D Imaging Radar also provides high-resolution imaging for elevation.
This can provide crucial information about objects on the road and determine if they need to be avoided or if they are part of the environment. Using 4D Imaging Radar, a sensor can identify key objects or obstructions (such as a pothole or a bridge), even while covered in a thick layer of fog, and safely direct the car on the best path.
Radar with Perception
Core functions such as SLAM or free space mapping heavily depend on the use of optical sensors; if extreme weather affects these sensors, then a vehicle needs to have redundancy to maintain safety. Unfortunately, most radar systems are not equipped to provide the level of redundancy required.
4D Imaging Radar can provide detailed information by use of its ultra-high resolution and state-of-the-art post-processing which includes:
- Real-time calculations of ego-motion
- Tracking and separating objects to fully understand the car’s surroundings
- Mapping of stationary objects
- Conducting Free Space Mapping
It is integral that Level 2 and above autonomous systems have these capabilities implemented in a reliable and safe manner., especially in the winter. 4D Imaging Radar provides an unparalleled level of safety in all weather conditions and environments enabling the safety measures that autonomous vehicles need to operate year-round.
Kobi Marenko (pictured, above left) is CEO of Arbe, a provider of next-generation 4D imaging radar chipset solutions.
