Bosch, TomTom collaborate on radar localization layer for high-resolution maps for autonomous driving
Bosch and the Dutch map and traffic information provider TomTom have developed a radar localization layer for high-resolution maps for automated driving—a first. Up to now, video data has been used for this purpose.
Bosch’s “radar road signature” is made up of billions of individual reflection points. These are formed everywhere that radar signals hit—for example, on crash barriers or road signs—and reproduce the course a road takes. Automated vehicles can use the map to determine their exact location in a lane down to a few centimeters.
The advantage of Bosch’s radar road signature is its robustness: unlike maps that rely exclusively on video data for vehicle localization, the radar road signature also works reliably at night and in conditions of poor visibility. Moreover, the Bosch radar road signature only transmits five kilobytes of data to a cloud per kilometer—this data volume is twice as high when using a video map. Bosch expects the first vehicles using the technology by 2020 at the latest in Europe and the US.
The two companies have been working intensively on the radar road signature and its integration into TomTom’s high-resolution overall map since the beginning of their collaboration in July 2015.
Bosch is a global market leader in the field of radar sensors with 77 GHz technology and detection ranges of up to 250 meters. By comparison, video sensors only have a maximum detection range of 150 meters. The main challenge was finding a way to adapt existing radar sensors for this task.
When used in a driver assistance system such as automatic emergency braking systems or adaptive cruise control (ACC), the sensors detect moving objects. But in order to generate the radar road signature, they also need to be able to detect static objects, which meant that existing radar sensors had to be modified. The next generation of Bosch radar sensors will be able to provide the data required for the radar road signature.
High-resolution maps are essential for automated driving and provide information that goes beyond the area that sensors are able to monitor. Unlike maps for today’s navigation devices, they are made up of a number of overlapping layers:
Localization layer: An automated vehicle can determine its position in a lane by using a localization layer comprising the Bosch radar road signature plus an additional video localization map. It compares information about objects that it has received via its surround sensors using the corresponding information from the localization layer. In this way, the vehicle can determine its relative position to these objects.
Planning layer: The planning layer is used to calculate individual maneuvers during automated driving (trajectory planning). The planning layer also contains information about the course of the road, traffic signs and speed limits, as well as bends and gradients. An automated vehicle can use the planning layer, for example, to decide when it should change lanes.
Dynamic layer: Information about any rapidly changing traffic situations such as traffic jams, construction work and hazards, or available parking spaces is saved in the dynamic layer.
Each of these layers of a high-resolution map for automated driving need to be regularly updated—the dynamic layer even in real time.
We currently expect that we will need fleets for freeways in Europe, North America, and Asia Pacific that each consist of around one million vehicles in order to keep our high-resolution map up to date.—Bosch board of management member Dr. Dirk Hoheisel
Current data for each of the layers will be generated by the vehicles’ on-board sensors while they are driving. Communication boxes such as Bosch’s Connectivity Control Unit will transmit the radar sensors’ data from the vehicles to the manufacturers’ cloud and then on to the Bosch IoT Cloud. Bosch will use this to create the radar road signature, which is compatible with all conventional map formats. TomTom’s responsibilities will include integrating the radar road signature in the overall map and distributing it.