Project Details
Projekt
Autonomous research vehicle
Subject Area
Systems Engineering
Term
Funded in 2021
Project identifier
Deutsche Forschungsgemeinschaft (DFG) - Project number 459365611
In order to significantly advance the development of algorithms for autonomous driving, they must be tested both in simulation with a large amount of data and on real hardware. The large-scale equipment applied for, will enable the Technical University of Munich for the first time to evaluate algorithms from simulation to real tests and to make the algorithms and recorded measurement data available to a large user community afterwards. By the joint effort of 10 professorships (7 computer science, 1 electrical engineering, 1 mechanical engineering, 1 traffic engineering), we expect a high utilization of the requested device.The presented utilization concept provides that the degree of maturity of the newly developed algorithms from the individual research groups is first determined automatically with the help of a Hardware-in-the-Loop Simulator (HiL) on real development hardware similar to the hardware in the test vehicle. In this first step, statements about the behavior and performance of the algorithms in the real vehicle can be made. Only when a certain degree of maturity is reached, these algorithms will be tested on the test vehicle in the second stage. Under the supervision of trained safety drivers, motion planning algorithms will be tested extensively in real traffic. In addition, the test vehicle can be driven by a human driver to record test data, evaluate algorithms for sensors and communication systems or to perform studies regarding human factors. During each test drive, all resulting data is recorded and transferred to the requested data center in the third stage after completion of the test drive. There, the data is post-processed and made freely available online to the entire research community. These data can then be used by computing clusters for machine learning. Furthermore, the data serve as new test cases for the HiL simulations and as a basis for the development of new algorithms by the applicants. Our usage concept thus allows to increase the maturity of the algorithms over time.A special focus of this application is to allow researchers worldwide to participate in the results. In order to make the data available in a prepared form, special emphasis is placed on the processing of data and its structured preparation. Since many research groups do not have access to data on autonomous driving, our project will also enable many other research groups to conduct top-level research in this promising area.
DFG Programme
Major Research Instrumentation
Major Instrumentation
Autonomes Forschungsfahrzeug
Instrumentation Group
2580 Zubehör und Aggregate für Kraftfahrzeuge
Applicant Institution
Technische Universität München (TUM)
