95 Transportation Research Record: Journal of the Transportation Research Board, No. 2316, Transportation Research Board of the National Academies, Washington, D.C., 2012, pp. 95–105. DOI: 10.3141/2316-11 A. Abdelghany, College of Business, Embry-Riddle Aeronautical University, 600 South Clyde Morris Boulevard, Daytona Beach, FL 32114. K. Abdelghany, Department of Civil and Environmental Engineering, Southern Methodist Univer- sity, P.O. Box 750340, Dallas, TX 75275-0340. H. S. Mahmassani, Transporta- tion Center, Northwestern University, 215 Chambers Hall, 600 Foster Street, Evanston, IL 60208. A. Al-Zahrani, Civil Engineering Department, King Abdul- aziz University, P.O. Box 80204, Jeddah 21589, Saudi Arabia. Corresponding author: A. Abdelghany, ahmed.abdelghany@erau.edu. tools are expected to provide architects and planners the appropriate platform for designing pedestrian facilities and developing effective schemes for their management. This paper presents an approach intended to fill the gap. It consists of a mesoscopic simulation-based dynamic trip assignment model, called PEDSTREAM, for studying crowd dynamics in networks of congested facilities. The model represents congestion dynamics that result from loading a given pedestrian time-varying demand pattern, as a function of available capacity along the different walkways and implemented crowd management strategies. The model can con- figure the study area in the form of a network and represent pedes- trian demand at the individual level. Given the pedestrians’ planned activities in the area, each pedestrian is assumed to choose a route that minimizes her or his travel disutility (e.g., travel time). Pedestrian movements along their routes are simulated by a mesoscopic relation- ship that represents the average walking speed as a function of the surrounding concentration. The model allows tracking of pedestrian movements in time and space and provides the capability for iden- tifying congested sections with potential flow breakdown. The list of the most efficient routes between every origin-destination pair in the area is updated periodically to reflect congestion dynamics in the area. The model produces an array of performance measures and provides animation capabilities that help in comparisons of the effectiveness of design alternatives and control schemes. The flow modeling features and network performance evaluation capabilities of the model are illustrated through application to a major pedestrian corridor in Makkah, Saudi Arabia. The model is used to develop a simulation-assignment platform for the Al-Mashaa’er corridor, which serves about 3 million pilgrims in a peak period that lasts several hours. The platform provides a decision support envi- ronment for the evaluation of design alternatives under anticipated future pedestrian demand levels and proposed crowd management strategies. The next section reviews related studies with a concentration on network-based pedestrian models. A description of PEDSTREAM’s modeling approach is then presented, followed by the application of the model for the Al-Mashaa’er corridor in Makkah, Saudi Arabia. BACKGROUND Considerable research has been performed in the past few decades to model pedestrian flow pattern in crowded facilities. The litera- ture reports numerous models that differ in their application scope, underlying theoretical approach, required effort for calibration and Dynamic Simulation Assignment Model for Pedestrian Movements in Crowded Networks Ahmed Abdelghany, Khaled Abdelghany, Hani S. Mahmassani, and Abdulrahem Al-Zahrani The planning of crowded pedestrian facilities involves providing adequate capacity for all walkways, as well as designing effective schemes for crowd management. PEDSTREAM is a mesoscopic simulation-based dynamic trip assignment model for large-scale pedestrian networks. The model can represent temporospatial distribution of pedestrians and associated service levels over the network and can predict pedestrian responses to changes in design, operational conditions, and crowd management strategies. The model was used to develop a simulation platform for the Al-Mashaa’er corridor in Makkah, Saudi Arabia, which served about 3 million pilgrims in a peak period that lasted several hours. Use of the plat- form is illustrated for the evaluation of design alternatives under future pedestrian demand levels and proposed crowd management strategies. Plans for crowded pedestrian areas such as city centers, tourist attractions, university campuses, stadium zones, and worship sites require adequate capacity for all planned walkways, as well as effec- tive schemes for crowd management. These needs require accurate representation of the pedestrians’ temporospatial distribution and asso- ciated service levels within the area. In addition, pedestrian responses to changes in design, operational conditions (i.e., incidents), and pro- posed crowd management strategies must be predictable. Although considerable research has been devoted to modeling pedestrian movements in crowded conditions, most models have focused on the microscopic behavior of pedestrians along predefined routes. Consequently, the models have intensive computation requirements that preclude real-world application to large-scale facilities or systems with high demand levels. Most models fall short of representing the dynamics of pedestrian route choice, which is an essential capabil- ity, especially in extremely crowded environments with alternative routing options. Thus, pedestrian modeling tools that can represent pedestrians’ global temporospatial distribution while being com- putationally optimized for large-scale applications are needed. Such