The increasing use of automated driving functions requires continuous monitoring of the immediate vehicle environment. Automated Valet Parking (AVP) is one automated driving function that is already in use due to the easy-to-control environment. The application is currently still in a highly controlled environment (no external disturbances by other road users and known environment) and depends on available infrastructural information. With regard to the implementation in mixed traffic and thus the use of all parking areas for the application of AVP, it is necessary to localize oneself in an unknown environment and to react to e.g. suddenly appearing pedestrians, other vehicles or obstacles within the driving trajectory.
In order to enable an unrestricted use of the AVP, algorithms for environment detection and localization, test methods and new sensor concepts are being developed in cooperation with Expleo GmbH and ASMPT within the Safe2P project. On the part of the Technical University, a special focus is on the detection and prediction of pedestrians using radar and camera sensors. Pedestrians represent a particularly complex and vulnerable road user in the environment detection due to their high flexibility of movement due to low restrictions and at the same time low protection possibilities. Test methods and objects, such as dummy pedestrians, are also being further developed to test the algorithms. The CARISSMA indoor test hall offers the possibility to simulate different scenarios for training and testing the algorithms.
Duration: 01.12.2021 – 30.11.2024
Funding code: DIK-2108-0009
Funding organisation: Bayerisches Staatsministerium für Wirtschaft, Landesentwicklung und Energie