RI: Small: Scalable Roadmap-Based Methods for Simulating and Controlling Behaviors of Interacting Groups: from Robot Swarms to Crowd Control

点击翻译按钮获取中文摘要

Simulating and controlling communities of characters that can interact with each other and their environment, and dynamically react to changes, is a challenging problem with many important applications ranging from homeland security (e.g., simulation of disaster scenarios and responses), to civil crowd control (e.g., planning exit strategies for sporting events), to education and training (e.g., providing immersive museum exhibits and training systems). While there are existing methods that attempt to address the simulation aspect, there is a lack of methods that support interaction of multiple types/groups of agents and little work has been done on the control or steering aspect. This work aims to address these challenges by integrating roadmap-based planning with agent-based modeling. This hybrid approach enables the development of methodology for modeling group interactions which are also influenced by constraints imposed by the environment (e.g., wide or narrow corridors) and techniques, including interfaces that enable planning and experimentation, that can scale to large numbers of agents. The results of this work will be shared with the community via publications and open source software. An anticipated outcome of this research is a tool for simulation and control of large crowds at major events (e.g., sporting events, political rallies, emergency evacuations of a building, region, or city). This could allow emergency response planners to investigate the crowd response when officials are placed in particular positions, or architects to study how evacuation times are affected by widening or narrowing corridors.