Robert Bosch, a German auto parts manufacturer, and TomTom, a Dutch mapmaker, are currently developing a high-resolution mapping system which employs Bosch radar sensors.
This elaborate technology will be installed in autonomous cars, thus facilitating the vehicles’ navigation by pinpointing their precise location. As of this day, video streaming is usually used for this purpose, but with this new technology, road data will be collected in real time using the radars, which have a larger range and detail compared to video cameras streaming. The technology implies using multiple radar signal reflection points which will help to reproduce real roads on maps. ‘Radar road signature’, as Bosch calls this technology, TomTom will integrate into its mapping system.
The radar road signature will also be more reliable than video data as it also operates perfectly well in low visibility condition or at night. One more advantage of this novice radar technology speaks for itself: using it, as little as 5 kB of data collected per each kilometer will be transmitted to cloud storage, and that is twice lower than with a video mapping system.
Dirk Hoheisel, a Bosch representative, stated that transportation means on the future market are going to run along roads utilizing a special sophisticated map created for driverless cars.
First vehicles aimed at collecting data for the Bosch’s radar road signature are expected on the roads the U.S. and in European countries. All in all, a 1 million vehicle fleet will operate on the world roads to make accurate and up to date maps.
The two enterprises have been cooperating since 2015, trying to integrate Bosch’s radar technology into high-resolution map created by TomTom. Originally, Bosch developed its radars for driver assistance functions. 77-GHz radar technology detects objects in motion at a distance of about 270 yards. That’s 105 yards further compared to camera sensors. But the technology needs to be adapted to also detect non-moving objects and make maps more holistic.
Autonomous cars cannot do without high-definition maps because they provide data which radar sensors alone aren’t able to trace. In fact, such maps consist of layers which overlap each other. It is necessary to constantly update these layers to be used in self-driving vehicles. So, massive amounts of data have to be constantly collected.
Currently, many well-known map providers are planning to use data collected from sensors on self-driving cars in the nearest future to build 3D maps in real time.