Workshop on pedestrian models 2014


Tentative schedule

14:1014:40Ulrich Kemloh
14:4015:10Guy Cooper
15:2515:55Frederik Johansson
15:5516:25Marija Nikolic
16:4017:10Toshaki Hiroyoshi
17:1017:40Sofia Kalakou
9:3010:00Serge Hoogendoorn
10:0010:30Flurin Haenseler
10:4511:15Jun Zhang
11:1511:45Antonin Danalet
12:0012:50Seminar by Alexandre Torday (Room GC C3 30)
14:2015:50Workshop discussions




12 presentations ordered by alphabetical order of the presenter (click on the title to access the slides, when available).


Cooper Guy, EPFL
A multi-class framework for a pedestrian cell transmission model accounting for population heterogeneity
A macroscopic cell transmission model for simulating pedestrian flows and congestion in a transportation hub is extended to account for pedestrian heterogeneity using a multiclass approach. The pedestrians are partitioned into sub-populations, or classes, wherein the behavior of pedestrians of the same class is homogenous across the sub-population and characterized by a fundamental diagram distinct to that pedestrian class. The network demand and affiliation of pedestrians to the various sub-populations are presumed to be known a priori. The derivation and inherent assumptions of this numerical model are presented in detail.
Danalet Antonin, EPFL
WiFi-Based Marauder's Map, or where are members of a campus and why?
Similarly to the Marauder's Map in Harry Potter, we tracked employees and students on campus. We model this output in terms of choice of activity type in time. By discretizing time, this choice is represented as a path in a network. The choice set is generated using a Metropolis-Hasting algorithm based on attractivity measures, and the utility function describes satiation effects and time-of-day preferences. Such a methodology is useful to build an activity-based model for pedestrian infrastructure, estimate the demand and its drivers, and forecast it for different scenarios.
Haenseler Flurin, EPFL
Schedule-based pedestrian demand estimation in a quasi-uncongested railway station
A framework is presented for estimating pedestrian demand within a railway station based on historical data, the train time table and link flow counts. The problem is considered in discrete time and at the aggregate level, i.e., for groups of pedestrians embarked on the same route and with the same departure time interval. For flow assignment, an empirical speed distribution is assumed.
Hiroyoshi Tosaki, Univesité Paris-Est
A route choice model of commuters in French railway stations, based on a discrete choice model
In the French railway stations which have become increasingly complex especially in Paris, we need prediction methods of pedestrian flows and we are developing some models of station user behaviors in order to make space sizing efficient. In the presentation, a route choice model based on a discrete choice model and calibrated with empirical data of Paris Saint-Lazare station provided by the French National Railway Company (SNCF) will be shown. This simple model assumes that passengers are homogeneous and know alternative routes at stations. Furthermore it will be the base model to introduce others factors influencing such heterogeneous passengers, congestion in the corridors and escalators.
Hoogendoorn S.P., Delft University of Technology
Pedestrian Flow Modelling: From data to models, from micro to macro...
Johansson Fredrik, Linköping University
Local performance measures of pedestrian traffic
A set of local performance measures of pedestrian traffic is presented. The specifications of the measures are motivated by the structure of the Social Force Model (SFM), and are based on a simple Gaussian filter. The measures are continuous in space and provides information on a spatial scale interesting for evaluation and design purposes. An analysis of some simple simulated scenarios using the measures are presented, providing insight into potential problems concerning the connection between the SFM and the route choice model. Further analysis of this problem is presented and a possible path to determining the severity of the problem is sketched.
Kalakou Sofia, Instituto Superior Tecnico - École polytechnique fédérale de Lausanne
A pedestrian simulation model for airport performance analysis
Transport terminals often face capacity limitations and their managers are challenged to deal with them. In such an attempt, simulation tools have been widely used to analyze terminal performance. Here, a methodology is presented for the simulation of pedestrian movements and passenger-related processes in an airport building. The developed model simulated the movements of passengers in an airport and the processes they pass through since they enter the airport building until they finish the security control process. Aspects related to the walking behavior and pedestrian group behaviors were included in the model and the results provided insights for the provided LOS at the airport.
Kemloh Ulrich, Forschungszentrum Jülich GmbH
Conception, development, installation and evaluation of a real time evacuation assistant for complex buildings
In this presentation a pedestrian evacuation system for large buildings is presented. The aim of the system is to monitor and forecast pedestrian streams in complex facilities to assist decision makers and security services in case of emergency. Since the last decades the study of pedestrian dynamics has gained more attention due to continuously growing urban population and cities. This growth is accompanied by an increase not only in size but also in frequencies of large-scale events. The increase sets new challenges to architects, urban planners and organizers of such events. The growing interest in large-scale events additionally requires increased security measures and new security concepts that are tailored to the large amount of people. The hereby designed system is capable of performing real time forecasts of pedestrian traffic for the next 15 minutes starting from the current situation. The test venue for the system is a multi-functional arena in Düsseldorf, North Rhine-Westphalia, Germany. The emphasis in this contribution is set not only on the development and the deployment of the system but also on the validation of the results during the operational phase.
Nikolic Marija, EPFL
Probabilistic speed-density relationship for pedestrians based on data driven space and time representation
In this presentation a framework that provides the detailed characterization of the pedestrian flow will be shown. The key components of the proposed framework are: (i) data driven space discretization based on the Voronoi tessellations that allows pedestrian oriented definition of density indicator; (ii) statistical and data driven approach to time aggregation, allowing for the pedestrian oriented definition of speed indicator; (iii) probabilistic model for speed-density relationship so as to capture the empirically observed heterogeneity among pedestrians. The calibration and validation results based on a real case study will be presented and analyzed in details.,
Zhang Jun, Forschungszentrum Jülich GmbH
Empirical characteristics on different types of pedestrian flows
In this presentation, series of well controlled experiments including (uni- and bidirectional flow as well as crossing flow) under laboratory conditions will be shown. To resolve the fine structure of the resulting velocity-density relations, fundamental diagram, the Voronoi method is used to analyze these experiments. The similarities and difference among these streams, which are necessary for fire safety design as well as pedestrian model calibration, will be presented.


April 10 - 12, 2014


EPFL Room GC D0 386


CHF 200


Mila Bender

Transport and Mobility Laboratory (TRANSP-OR)
Station 18
CH-1015 Lausanne

Tel: +41 (0) 21 693 24 08
Fax: +41 (0) 21 693 80 60