MSc thesis project proposal

Radar based Road Mapping and Vehicle Localization

Project outside the university

TomTom

Highly accurate localization of moving vehicles in urban environments is crucial for emerging autonomous driving. Current approaches integrate global positioning system (GPS), inertial measurement unit (IMU), wheel tick odometry, high-resolution panoramic imagery and light detection and ranging (LIDAR) data acquired by an instrumented vehicle, to generate high-resolution environment maps that is used for localization. Most of these sensors are not yet part of the standard equipment for modern cars. Therefore gathering the data for these maps still requires expensive special vehicles. Due to this, the process of gathering data for map generation is expensive and time consuming.

Since radar sensors are already available in most of the modern cars, in this master thesis project, we design a method/algorithm that uses radar sensor to create road maps that can be used for localization.

The proposed project has the goal to develop a prototype demonstrating the feasibility to create a radar map that can be used for localization based on the output (peek-lists) from off-the-shelf radar sensors.

Assignment

The list underneath provides an overview of the research questions to be answered in this project:

  • Are radar measurements of the same sensor under similar circumstances repeatable
  • Are radar measurements from different radar sensors repeatable
  • Can the aggregated radar measurements be correlated with TomTom’s LIDAR point cloud data

Answers to the above questions should provide sufficient basis to draw conclusions about the feasibility to create a radar map layer from LIDAR data.

Contact

dr. Faruk Uysal

Microwave Sensing, Signals and Systems Group

Department of Microelectronics

Last modified: 2020-09-23