Artificial Intelligence in Traffic Engineering and Control

In the Focus Group Artificial Intelligence in Traffic Engineering and Control, Rudolf Diesel Industry Fellow
Felix Rempe (Autonomous Driving, BMW Group) collaborates with his host Prof. Klaus Bodenberger (Chair of Traffic Engineering and Control, TUM).

The development of connected and automated vehicles promises a profound change of current transportation systems. Using multiple sensors and fast transmission of data, significantly more information about the current traffic state, including all road users, can be gathered. Additionally, and most important, in an automated vehicle the human driver will be assisted or replaced with a computer system that acts in a deterministic and controllable way. The vehicle’s speed and trajectory can be predicted and altered if necessary. Thus, foreseeable conflicts between road users can be avoided preemptively, inefficient stop and go movements can be suppressed and road space can be exploited better.

Within this Focus Group novel approaches for traffic state estimation and control exploiting the technical opportunities of connected and automated vehicles are developed. Therefore, large amounts of data collected by current vehicle fleets are used to train machine learning algorithms to accurately infer current and future traffic conditions in transportation networks. On top of these, traffic control algorithms are designed, implemented and validated using the TUM testbed for automated driving and simulation environments.