Modeling and Simulation of Crowd Behavior

This project presents the simulation and modeling of a crowd behavior in a virtual environment. Many research and social interaction aspects of crowd behavior are beyond the reach of available tools. The VR environment is able to adapt for specific design situations and provide insight to current perspective as well as what-if questions. Our hypothesis is that mutual concern, helping, and co-ordination will more likely occur when crowd members share a common identity. The selfish behavior associated with 'mass panic' may occur in evacuation if there is no shared identity. Our hypothesis is that the visualization studies support the conclusion that, where there is a strong sense of collective identity, there will be mutual concern and helping.

View of library Shelves and Avatars Agents moving towards the goal in library

This project presents the simulation and modeling of a Bowie State University library building in a virtual environment. The virtual library building is modeled on a real time library building on campus at Bowie State University. Depending on the distance between the agent and the target, different behaviors will gets assigned to the respective agent. Finally, the seek behavior takes an agent to the target. The transition lines includes the max script to change the states of an agent.


During emergency evacuation, behavior of people becomes uncertain. Each agent’s emotions such as stress, anger, and panic are defined as low, medium, and high. Rules are intuitively designed based on each agent’s location and speed.

Simulating Behaviors for various agents using cognitive controller in MATLAB Behavior List for avatars in 3ds Max

Embedded Video

Embedded video with a virtual environment can be used to show multiple views for an evacuation visualization task.Embedded video visualizations integrate video and 3D model data to help users understand complex situations. Fig. 5show an agent visualizing the embedded video for exit routes inside the building during evacuation. When evacuating the building, the user relies more on the embedded videos. We plan to further explore several designs from the video processing dimension and the navigation dimension as well in our future work.

Fig.4. View of an avatar visualizing the embedded video during evacuation.


  • Sharma, S., Vadali, H., "Modeling Emergency Scenarios in Virtual Evacuation Environment", IEEE World Congress on Computer Science and Information Engineering, CSIE, March 31 - April 2, Los Angeles, USA,2009
  • Sharma, S., Vadali, H., “Simulation and Modeling of a Virtual Library for Navigation and Evacuation”, MSV'08 - The 2008 International Conference on Modeling, Simulation and Visualization Methods, Monte Carlo Resort, Las Vegas, Nevada, USA, July 14-17, 2008
  • Sharma, S., Vadali, H., “Fuzzy Logic Application for Emergency Evacuation in a Virtual Environment”, Proceedings of World Multiconference on Systemics, Cybernetics and Informatics, WMSCI-2006, Orlando, USA, July, 2008 (accepted)