Automated construction of variable density navigable networks in a 3D indoor environment for emergency response

摘要

Widespread human-induced or natural threats on buildings and their users have made preparedness and quick response as crucial issues for saving human lives. Available information about an emergency scene, e.g. the building structure, material and trapped people helps for decision-making and organizing rescue operations. The ability to evaluate potential scenarios for human evacuation, and then identifying the paths of egress during an emergency is critical for rescue and emergency services. Good quality models supporting real, or near-real, time decision-making and allowing the implementation of automated methods are highly desirable. In this paper, we propose a new automated method for deriving a navigable network in a 3D indoor environment, including a full 3D topological model which may be used not only for standard navigation but also for finding alternative egress routes and simulating phenomena associated with disasters such as fire spread and heat transfer.