Abstract
Abstract. Egress (or Evacuation) modelling has the purpose to simulate how the people inside a building can move outside in a safe manner and as quick as possible in case of an emergency. The goal can be to assess the safety of an existing building or a building design, to establish evacuation routes and install signage, or even to give real-time guidance to the evacuating crowd during an emergency. In all cases, the availability of a three-dimensional building model allows to generate the necessary input for the egress modelling software. The algorithms for such software can be subdivided into two major categories, working with navigation grids or navigation networks, respectively. The purpose of this paper is to illustrate that voxel-based building models are particularly suited to provide input for both categories. We further illustrate how to process might continue using an algorithm based on (electrical) network analysis, which can be applied in both a gridded and a network-based simulation. We provide preliminary results for the gridded case.
Highlights
At the Geospatial Research, Innovation and Development (GRID) group at the University of New South Wales we are developing a comprehensive infrastructure for the representation of three-dimensional geographic information using voxels, next to the more common vector representation models
ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Volume IV-4/W9, 2019 4th International Conference on Smart Data and Smart Cities, 1–3 October 2019, Kuala Lumpur, Malaysia illustrate our findings, we introduce an evacuation modelling approach based on Kirchhoff’s first law for electrical circuits, which can be applied to gridded and networks models alike
We have demonstrated that a 3d voxel building model can be a good basis for studying various aspects of Egress modelling, using different approaches
Summary
At the Geospatial Research, Innovation and Development (GRID) group at the University of New South Wales we are developing a comprehensive infrastructure for the representation of three-dimensional geographic information using voxels, next to the more common vector representation models. Even when the crowd behaves ‘well’, each individual cannot predict to which extent doing ‘the right thing’ may lead to congestion later in the process Such a prediction would require that everybody has full knowledge about everybody else’s current position and future behaviour, and of the full consequences of planned and alternative behaviours. The purpose of the current paper is mainly to show that both approaches (gridded and network) can have their inputs derived from a voxel building model. We feel that the gridded approach (a) may be the more suitable one in general and (b) fits better to the voxel representation of our input model – this formulation is meant to leave room for discussion and is not intended as a conclusion or a verdict. We introduce both approaches but work out the gridded approach in greater detail
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
