Cooperative and Intrinsically-Correct Control of Vehicles in Diverse Environments (CoInCiDE)
Final Report Abstract
The self-driving car will be more efficient with resources and safer for occupants and other road users than today’s vehicles. These and other advantages associated with automated driving can only be fully achieved and exploited through cooperation between road users. The DFG-funded project Cooperative and Intrinsically-Correct Control of Vehicles in Diverse Environments aims to contribute to this goal. For this, the project explored how the negotiation of safe cooperative maneuvers between automated vehicles can be ensured using vehicle-to-vehicle communication, how cooperative behavior in emergency situations can contribute to safe motion planning, and how decision-making functions can be designed for automated cooperative vehicles. To efficiently investigate the aforementioned issues, the open-source software Eclipse-ADORe was used and extended. This allows the simulation of the various concepts and contains interfaces to other standard frameworks for research on automated driving. ADORe also allows the operation of the DLR research vehicles, which enable the testing of the respective implemented algorithms and methods in a practice-relevant system environment. With the Space-Time-Reservation-Procedure (STRP), a method was developed and researched that enables the negotiation and execution of safe cooperative maneuvers. In this context, various key requirements were successfully addressed: First, universal applicability is elementarily necessary, through which the use is not limited to specific traffic situations. Furthermore, the STRP must be resilient to disruptive factors such as the failure of the communication link and the possibly resulting loss of messages. In addition, the method is suitable for use in mixed traffic, i.e., traffic in which manually operated or automated vehicles of different automation levels also participate. The method has been successfully tested in simulation and test drives. In addition, decision-making for automated and cooperative vehicles was investigated. Here, effective decision-making functions could be built using various methods, including machine learning. Furthermore, maneuver-level cooperation for handling emergency situations was investigated. In an example scenario, the area directly in front of the automated vehicle is suddenly blocked, and a lane change is initially not possible due to another vehicle. In some situations, braking to standstill in front of the obstacle cannot be realized due to physical limitations. For such cases and other cases where an evasive maneuver is more favorable, an approach for maneuver cooperation with negotiation via vehicle-to-vehicle communication in emergency situations was researched. It was shown in simulation that this can prevent collisions. Furthermore, the method was demonstrated at an international conference with the DLR research vehicles.
Publications
-
"Hierarchical approach for safety of multiple cooperating vehicles," Proc. of the AAET. ITS automotive nord e.V., March 2017
V. Jain; D. Heß; C. Löper; T. Frankiewicz & T. Hesse
-
"Safe cooperation of automated vehicles," Proc. of the AAET. ITS automotive nord e.V., March 2017
D. Heß, C. Löper & T. Hesse
-
Ensuring drivability of planned motions using formal methods. 2017 IEEE 20th International Conference on Intelligent Transportation Systems (ITSC) (c(2017, 10)). IEEE.
Schurmann, Bastian; Hes, Daniel; Eilbrecht, Jan; Stursberg, Olaf; Koster, Frank & Althoff, Matthias
-
Explicit Negotiation Method for Cooperative Automated Vehicles. 2019 IEEE International Conference on Vehicular Electronics and Safety (ICVES) (c(2019, 9)), 1-7. IEEE.
Nichting, Matthias; Hes, Daniel; Schindler, Julian; Hesse, Tobias & Koster, Frank
-
Negotiation of Cooperative Maneuvers for Automated Vehicles: Experimental Results. 2019 IEEE Intelligent Transportation Systems Conference (ITSC) (c(2019, 10)), 1545-1551. IEEE.
Hes, Daniel; Lattarulo, Ray; Perez, Joshue; Hesse, Tobias & Koster, Frank
-
Space Time Reservation Procedure (STRP) for V2X-Based Maneuver Coordination of Cooperative Automated Vehicles in Diverse Conflict Scenarios. 2020 IEEE Intelligent Vehicles Symposium (IV) (c(2020, 10, 19)), 502-509. IEEE.
Nichting, Matthias; Hess, Daniel; Schindler, Julian; Hesse, Tobias & Koster, Frank
-
"Infrastructure-aided Automated Driving in Highly Dynamic Urban Environments," ITS World Congress 2021, Hamburg, Deutschland
S. Lapoehn; D. Heß; C. Böker; H. Böhme & J. Schindler
-
Case Study on Gap Selection for Automated Vehicles Based on Deep Q-Learning. 2021 International Conference on Artificial Intelligence and Computer Science Technology (ICAICST) (c(2021, 6, 29)), 252-257. IEEE.
Nichting, Matthias; Lobig, Thomas & Koster, Frank
