Study of Human Factors Contributing to Fatal Injury in a Multi-floor Building during Emergency Using Finite State Machines

Abstract

In this paper a finite state machine model for human behavioral tendencies under emergency situations in multi-floor buildings is proposed. Prediction of human activity inside high rise buildings on emergency scenarios like earthquakes, tsunami, fire accident or terrorist attacks is a difficult task which could cause fatal injuries and needs realistic modelling of the scenario. Finite state machines which could be analyzed using Markov chains facilitates the analysis of human behavior. We develop a finite state machine model that captures human behavioral state characteristics and provides the solution based on discrete-time Markov chain. We extend our method to the specific case of building with multiple floors and stairways to analyze the behavior of occupants under an emergency condition and evaluate the overall evacuation time.