Realizing a Demonstrator for the Autonomous Exploration of a Parking Lot
by , , , , , , , , , , ,
Abstract:
Driver assistance systems in modern vehicles are becoming increasingly widespread and their quality is constantly improving. Nevertheless, a big step is necessary to progress from these systems to fully autonomous vehicles that make sensible decisions in every situation. A reasonable intermediate step on this path is the implementation of self-driving cars that are specialized for specific tasks. In this context, we present a demonstrator for the driverless exploration of a parking lot. Particularly we give details about the hardware configuration and present the key ideas of the algorithms used. The latter emerge from an modular concept consisting of two main steps: (i) multi-sensor fusion and (ii) model predictive control. The proposed demonstrator can navigate autonomously on a parking lot and dynamically take obstacles into account. Free parking spaces are furthermore identified and appropriate parking maneuvers are performed.
Reference:
Realizing a Demonstrator for the Autonomous Exploration of a Parking Lot (Andreas Folkers, Joachim Clemens, Frank Schrödel, Matthias Rick, Arne Berger, Verena Schwarting, Constantin Wellhausen, Mitja Echim, Max Gerwien, Matthias Knauer, Kerstin Schill, Christof Büskens), In 1st Virtual IFAC World Congress 2020 (IFAC-V), 2020. (in press)
Bibtex Entry:
@inproceedings{folkers2020realizing,
	title = {Realizing a Demonstrator for the Autonomous
	Exploration of a Parking Lot},
	booktitle = {1st Virtual IFAC World Congress 2020 (IFAC-V)},
	year = {2020},
	note={in press},
	author = {Folkers, Andreas and Clemens, Joachim and Schr\"odel, Frank and Rick, Matthias and Berger, Arne and Schwarting, Verena and Wellhausen, Constantin and Echim, Mitja and Gerwien, Max and Knauer, Matthias and Schill, Kerstin and B\"uskens, Christof},
	abstract={Driver assistance systems in modern vehicles are becoming increasingly widespread and their quality is constantly improving. Nevertheless, a big step is necessary to progress from these systems to fully autonomous vehicles that make sensible decisions in every situation. A reasonable intermediate step on this path is the implementation of self-driving cars that are specialized for specific tasks. In this context, we present a demonstrator for the driverless exploration of a parking lot. Particularly we give details about the hardware configuration and present the key ideas of the algorithms used. The latter emerge from an modular concept consisting of two main steps: (i) multi-sensor fusion and (ii) model predictive control. The proposed demonstrator can navigate autonomously on a parking lot and dynamically take obstacles into account. Free parking spaces are furthermore identified and appropriate parking maneuvers are performed.},
}