Automotive -

HD Map Localization

 
 
 

The SLAM Challenge

HD Maps contain a trove of information on roads and roadside objects such as road topology, lane markings, road barriers and signs, which are used for autonomous driving and various ADAS features. In order to extract such information from HD Maps, GPS or GNSS is used to position within these maps. More advanced systems also fuse IMU or rely on LiDAR, while mass production passenger vehicles are matching street signs or lane markings to the HD Map. None of these can reliably maintain the required positioning accuracy in the HD Map to grab the required information.

How ArtiSLAM Can Be Applied

ArtiSLAM provides a solution to enable localization within HD Maps in any environment. The combination of VIO and GNSS alone provides high accuracy global positioning, even during occasional outages of GNSS or poor GNSS quality. By additionally matching deep learned and actual road features to the existing map, and fusing with the VIO and GNSS information, accurate positions can be provided at all times.

How Clients Benefit

  • Extended coverage: Information within HD Maps are still accessible in GNSS-denied environments, even without LiDAR sensors or INS devices.
  • Additional redundancies: Even when GNSS is available, the VIO is always there as a second system to cross-check the integrity of the position against.
  • Greater safety & control: Being able to extract information from the HD Map in GNSS-denied environments ensures a more secure and controlled driving experience, whether used for self-driving or specific ADAS features.
 
 

With VIO, GNSS and additional matching of road features, ArtiSLAM allows automotive partners to localize and extract data within HD Maps across all environments.

 
 

Interested to apply ArtiSLAM to localize within HD Maps today? Let’s get in touch.