Understanding traffic situations in dynamic traffic environments is an essential requirement for autonomous driving. The prediction of the current traffic scene into the future is one of the main problems in this context. In this publication we focus on highway scenarios, where the maneuver space for traffic participants is limited to a small number of possible behavior classes. Even though there are many publications in the field of maneuver prediction, most of them set the focus on the classification problem, whether a certain maneuver is executed or not. We extend approaches which solve the classification problem of lane-change behavior by introducing the novel aspect of estimating a continuous distribution of possible trajectories.Our novel approach uses the probabilities which are assigned by a Random Decision Forest to each of the maneuvers lane following, lane change left and lane change right. Using measured data of a vehicle and the knowledge of the typical lateral movement of vehicles over time taken from realworlddata, we derive a Gaussian Mixture Regression method. For the final result we combine the predicted probability density functions of the regression method and the computed maneuver probabilities using a Mixture of Experts approach.In a large scale experiment on real world data collected on multiple test drives we trained and validated our prediction model and show the gained high prediction accuracy of the proposed method.

No abstract

Potholes and other damages of the road surface constitute a problem being as old as roads are. Still, potholes are widespread and affect the driving comfort of passengers as well as road safety. If one knew about the exact locations of potholes, it would be possible to repair them selectively or at least to warn drivers about them up to their repair. However, both scenarios require their detection and localization. For this purpose, we propose a crowd-based approach that enables as many of the vehicles already driving on our roads as possible to detect potholes and report them to a centralized back-end application. Whereas each single vehicle provides only limited and imprecise information, it is possible to determine these information more precisely when collecting them at a large scale. These more exact information may, for example, be used to warn following vehicles about potholes lying ahead to increase overall safety and comfort. In this work, this idea is examined and an offline executable version of the desired system is implemented. Additionally, the approach is evaluated with a large database of real-world sensor readings from a testing fleet and therefore its feasibility is proved. Our investigation shows that the suggested CPD approach is promising to bring customers a benefit by an improved driving comfort and higher road safety.