A newly developed radar system should allow autonomous vehicles to look around the corner. This allows vehicles, cyclists and pedestrians to be noticed in good time.
Dangers instead of speed
In principle, the system uses normal radar sensors, such as those used in speed cameras. However, the direct image is not important, as in the detection of razors. Instead, the approach uses signals that would be rejected as noise in other applications, namely radio waves that were reflected by surfaces such as house walls or even parked cars. These can ultimately fall onto a sensor on objects behind a corner and again via the wall.
In order to see around the corner with these actually weak signals of low resolution, specially developed algorithms are used that process the data with AI methods. «First we have to see if there is anything there. If there is anything, is it important? Is it a cyclist or a pedestrian? Then we have to localize it, ”says Fangyin Wei, a PhD student in Princeton. The system takes advantage of the fact that the Doppler effect occurs when radio waves are reflected from moving objects.