A 3DGIS multi-agent geo-simulation model for assessment of building evacuation scenarios considering urgency and knowledge of exits

ABSTRACT

A three-dimensional geographic information system (3DGIS)-based multi-agent geo-simulation model was developed using the GAMA simulation platform integrating 3D GIS layers and agent-based modelling for evacuation scenario modelling. The prototype model developed was used to generate building evacuation scenarios in order to perform a preliminary assessment of the effects of panic on the egress of building occupants, represented in the model as urgency (increased evacuation speed) and knowledge of exits. In the model, 3DGIS layers of the case study building were used as inputs to model the simulation agents referred to as ‘species’. PEOPLE species represents the building occupants evacuating to their target EXIT_POINTS along the INDOOR_PATHS. Their movement is affected by the chance of panic (CoP) and the chance of knowledge of exit (CoKE) parameters. CoP sets the number of PEOPLE agents ‘panicking’, which would be assigned with movement at 1.5 times the default speed. On the other hand, CoKE conditionally sets the target EXIT_POINT to either the designated exit attribute of the ROOMS where the PEOPLE species is located or to a random exit. Speed of movement along the INDOOR_PATHS is also affected according to path weights which are proportional to the number of PEOPLE species on the same path. Different simulation runs were performed, varying the CoP and CoKE. In general, knowledge of exit only reduced the total egress time slightly, but within the egress period, more people were able to exit quickly with increased knowledge of exit. Shorter egress time can be achieved with a random selection of exits and low CoP. However, as the number of occupants panicking increased, total egress time would eventually increase. The simulation illustrated the limitations of designating exits based on the shortest route only and indicate that knowledge of the volume of evacuating people along the indoor paths is beneficial for improving egress.

KEYWORDS:

• 3DGIS
• geosimulation
• building evacuation
• GAMA

Acknowledgements

Thanks are extended to Focus Global Inc. for the professorial chair award of the corresponding author. Likewise, the authors would like to acknowledge the indirect contribution of the GAMA developers and user community through their online forum at https://groups.google.com/forum/#!forum/gama-platform

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

E. G. Macatulad http://orcid.org/0000-0001-7977-2932

A. C. Blanco http://orcid.org/0000-0002-0489-9979

Additional information

Funding

This work was supported by the Department of Science and Technology Engineering Research and Development for Technology (DOST-ERDT) Programme.