Microscopic Jam Tail Warning for Automated Driving

Abstract

We perform a spatiotemporal differentiation between two different phases in congested traffic at the upstream front of the congested pattern: (1) the wide moving jam phase (J) and (2) synchronized flow phase (S) as introduced in the three-phase theory by Kerner. With this approach one can derive a conclusion how dangerous certain parts of the congested pattern really are and decide whether oncoming vehicles should be warned with a jam tail warning system. Each of the probe vehicles can distinguish either F→J or F→S transitions through a method for traffic phase identification. With the detailed information about the phase transitions happening at the upstream front of the congested pattern, the automated vehicle can drive safer and more comfortable while maintaining an unobtrusive behavior approaching traffic congestion. In our study made in this paper, empirical data is obtained through the use of probe vehicles with an average frequency of about 10 s. This allows us to reconstruct the F→J or F→S transitions over time at the upstream front of empirical congested patterns with a high quality that is sufficient for microscopic jam warning for automated vehicles.