PTV Viswalk: Quickest vs. Shortest Path in a Simulation of Pedestrians





Rating is available when the video has been rented.
This feature is not available right now. Please try again later.
Published on Jun 28, 2011

This animation demonstrates first the principle of how simulated pedestrians in PTV Vision's multi-agent crowd simulation VISWALK are made to walk along the path of estimated least remaining travel time under consideration of obstacles and other pedestrians. Second (beginning at 01:42) the effect of the method called "dynamic potential" is demonstrated by comparing it to pedestrians which are simulated such as if they would desire to walk along the shortest path (i.e. they minimize travel distance instead of travel time).

Much alike vehicle drivers, pedestrians desire to minimize their travel time to the destination; in some situations that desire is stronger than in others. Moreover, the walking direction for the quickest path is not easily to be determined. Details of the method how this determination is carried out will be published in an article in Advances in Complex Systems (see: http://dx.doi.org/10.1142/S0219525911...). A pre-print is available at the arXiv: http://arxiv.org/abs/1107.2004

Below follows a list of all episodes that are shown in this demo video - along with their start times. Please note that the various episodes are shown with different time lapse factors. The pedestrians in all situations move with approximately the same speed. (Note to everyone who has read the paper mentioned above: the impact strength of the dynamic potential was always set to 100%):

01:42 : This is a situation taken from real life: about 800 passengers alight from two trains arriving simultaneously at the station at the south entrance of Berlin's congress center (ICC). Making a large group of pedestrians walk realistically and efficiently around a sharp corner is one of the main applications of the dynamic potential method.
Note that with only a small group of pedestrians both trajectories would be very similar (quickest and shortest path are then almost identical).

03:18 : Now a large group of pedestrians has to take a U-turn. This is more difficult and therefore the difference between the two methods (left and right) is even more distinct.

04:48 : Here two large groups meet in a straight counterflow. This is a situation where the use of the dynamic potential does not necessarily produce a better result. However, it provides an alternative behavior that shows up after some seconds. The behavior on the left side is more realistic if the pedestrians assume that the counterflow situation will persist only shortly (e.g. at a busy traffic light), the right side is more realistic, if they assume that it will persist for longer (crowds moving through a city center festival).

06:18 : If counterflow occurs at a sharp corner, the dynamic potential (right side) is able to better reproduce that people move efficient in such a situation and most of them are able to resolve the situation (although at some high demand also in reality a crowd as seen on the left side will occur).

07:48 : Counterflow at a U-turn.

09:18 : A simple station hall. Some people (red) are urgently rushing for their train, some (green) have alighted from a train and have no reason to dwell in the station hall whereas some others (blue) have arrived at the station well before departure and now linger around. (Obviously the group is very large and behaves strangely, but it demonstrates the effect of the method better.) The red and green pedestrians in the upper left video follow the shortest path. With a growing blue group they get stuck in that group. The upper right and the lower situations were simulated with the quickest path approach but different values for parameter h (for details please refer to the paper linked above). Note that in the lower two videos the red and green groups manage to establish direction separation respectively spontaneous lane formation, while they do not for h=0 in the upper right example.

10:08 : This is an example from the laboratory. It does not even remotely occur in reality. However, it is a nice small example showing clearly and precisely the effect of the quickest path approach resp. dynamic potential.

10:48 : So far all route choice decisions have been continuous, i.e. the pedestrians could choose from a continuum of paths to their destination. This is the first example with a discrete route choice. The pedestrians need to choose if they walk the left or the right corridor. The method of the dynamic potential was not constructed for usage in such a situation. Other methods might be more helpful (in VISWALK for example partial routes), however, it is helpful as well here
12:43 : A grandstand. The interesting aspect of this example is that it includes effectively a sequence of one-dimensional segments (links) for the pedestrians. Therefore the direction of the shortest and the quickest path can differ by 180 degree. Here it is very obvious when pedestrians decide to walk a detour in exchange for a reduced travel time.


When autoplay is enabled, a suggested video will automatically play next.

Up next

to add this to Watch Later

Add to

Loading playlists...