> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 1 The Incorporation of Layout Design into an Agent-based Evacuation Model Tariq M. Elhassani, Jun Zhang, Mahmoud T. Khasawneh, and Ra’ed M. Jaradat to simulate the interactions between the three elements. Abstract—The innovative designs and layouts of buildings and public places have created a major issue in evacuating a massive group of people under intense and panic situations. In this paper, we developed an agent-based model to investigate the effect of layout design, people, and nature of threat on emergency evacuation. The purpose of this research was to provide a preliminary attempt at using agent-based modeling and simulation analysis to explore potential options that people have when faced with bomb explosion in an open market Environment. The results show that evacuation time, number of fatalities, and number of injuries are influenced by the layout design components, namely locations of rescue area, check point, and the bomb itself. Index Terms—Agent-based modeling, bomb explosion, emergency evacuation, layout design, marketplace, rescue operation. I. INTRODUCTION T HE ongoing trend of advanced knowledge in building designs and structures has raised major concerns for people safety. Innovative methods and approaches are needed to understand and assess these complex designs to assure people safety and verify compliance with standards and guidelines. Traditionally, verification has been demonstrated through full-scale evacuation drills and practices. A promising alternative to assess people safety lie's in computer evacuation models. The development of evacuation models in the last three decades has mainly contributed to the assessment of people safety and evacuation procedures in a variety of designs, under a range of environmental conditions. The effectiveness of such evaluation relies mainly on the models’ ability to reflect the detailed interactions between the occupant, building design, and environment. The evacuation models not only evaluate the efficiency of evacuation processes by controlling challenges presented in evacuation drills, but also simulate the dynamic interaction between the structure, people, and the environment. The purpose of this paper is to model the influence of structure layout, people behavior, and environmental conditions on the emergency evacuation process. An agent-based model has been developed T. M. Elhassani, J. Zhang, M. T. Khasawneh, and R. M. Jaradat are all graduate students in the Department of Engineering Management and Systems Engineering at Old Dominion University, Norfolk, VA 23529 USA (corresponding author’s phone : 757-683-5145; fax: 757-683-5640; e-mail: [email protected]).. II. LITERATURE REVIEW AND BACKGROUND Agent Based modeling (ABM) is classified as a computational model that simulates the actions and interactions of multiple autonomous decision-making entities [1]. Agent-based modeling, sometimes called individualoriented, or distributed artificial intelligence-based, is an increasingly powerful modeling technique for simulating individual interactions in a dynamic system and is distinctive in its ability to simulate the situation where the future is unpredictable [2]. ABM is used to develop models for a population dynamics system and a segregation system [1]. In the last few years, agent-based modeling has emerged in social, political, and economic science applications [3]. Agent-based modeling has played a key role in modeling human behavior and performance in emergency evacuations applications. Henein and White have proposed an efficient based model of crowd evacuation that incorporates pushing forces and injuries [4]. Their model was an enhancement of the pre-existing Kirchner model [5], and it investigates the force effects at different crowd densities [4]. Shendarkar, Vasudevan, Suengho, and Son have used virtual reality (BDI) to construct crowd simulation and demonstrate the use of the same for crowd evacuation management under terrorist bomb attacks in public areas [6]. The study was conducted to explore the crowd behavior under an evacuation situation in the event of a terror attack at the National Mall area in Washington DC [6]. The results obtained from this experiment gave an explicit insight into the patterns of human thought processes. Chen study aims to investigate the effectiveness of simultaneous and staged evacuation strategies in different road network structures by using ABM simulation [7]. The study was able to model the traffic flow at the level of individual vehicles in this case agents during an evacuation in natural way [7]. The results showed how the evacuation time changes in different flow modes. III. PROBLEM DESCRIPTION AND MODELING A. Problem Description Emergency evacuation is the urgent movement of people from dangerous place due to the threat or occurrence of a disastrous event [8]. Human life is always threatened by many hazards such as bombs, fires, explosions, and toxics. The > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < innovative designs and layouts of buildings and public places have created a major issue in evacuating a massive group of people under intense and panic situations. A mass movement of a crowd under panic can easily lead to injuries or deaths, if the crowd fails to run away from the hazardous area. In this paper, we developed an agent-based model to investigate the effect of layout design, people, and nature of threat on emergency evacuation. Our model represents an open market place under a bomb threat. It consists of a set number of shoppers and security officers placed randomly in the market, a bomb threat, a shelter place, and a check point. In addition to evacuation time, the model captures other output variables such as the number of fatalities and injuries, number of people who seek shelter at the public shelter place, and number of people who run toward the check point. B. Problem Modeling NetLogo is a cross-platform multi-agent programmable modeling environment [9]. NetLogo is one of the most used software in agent-based modeling simulation; it was originally developed in 1999 by Uri Wilensky, a professor of Learning Sciences and Computer Science at Northwestern University [10]. NetLogo is very simple software that enables one to understand the concept of agent-based modeling. NetLogo was initially used for educational purposes but later many researchers have found it a very easy and useful tool to learn and to conduct profound research. The proposed evacuation model in this paper shows the effect of layout design, people, and nature of threat on emergency evacuation. Our NetLogo, agent-based model (presented in Fig. 1) represents an open market place under bomb threats. It consists of agents (population), bomb (hazard), shelter, and check point. The inputs to the model include The location (coordinates) of the rescue area (white color). The location (coordinates) of the bomb threat. The displacement of the people (blue color). The displacement (locations) of security officers (orange color). The location (coordinates) of the check point. The location of the rescue area is assigned randomly. The purpose of that is to observe and study the relationship between its locations and the output variables. The rescue area is located anywhere in the model except the impact area of the bomb. There is no full capacity for the rescue area. When the bomb explodes, people who are closed to the rescue area will run to it. The location, impact and the explosion time of the bomb can be changed manually using the slide bars. The bombimpact ranges from 1 to 40. For the purpose of the study, we set the bomb-time at 10 ticks for all simulation runs. The bomb time indicates the time between the start of a simulation run and when the bomb explodes to guarantee the random movement of people in the marketplace. 2 The people in the model move in a random heading. There is no specific pattern or algorithm for the movement of people before the bomb goes off. Such setting makes the model more realistic to the problem at hand. Once the bomb goes off, people will move in the opposite direction from the bomb impact towards either the rescue area or the boundaries of the marketplace. Only the people who are in the rescue area’s influence distance will run away toward it. Otherwise they will evacuate to the boundaries. The influence distance, set at 14, identifies how close people are to the rescue area. It ranges from 0 to 20. There is no collusion between people during the evacuation. If the people trapped inside the bomb-impact zone, they will either die or injured. The “population” slide bar was used to change the number of the people. The number of the people ranges from 1 to 300. In all the runs we set the number at 150. Fig. 1. Snapshot of NetLogo before bomb deploys. The officers will also move in a random heading. Their primary task is to rescue the injured people once the bomb explodes. Accordingly, the code has been programmed to let the officers who are closed to the bomb-impact zone to run directly to rescue injures. The other officers who do not rescue the injured will not move (stop). The number of officers could range from 1 to 30. In this simulation we made sure that there is one officer for each injured person. The last input to the model is the location of the check point. It functions as a rescue area for the injured people. The non injured people will never escape to the check point. The location for the check point is fixed with all runs. We created a monitor to calculate the number of the injured in the check point. The evacuation time in this model starts once the bomb goes off until all injured people arrive at the check point. Evacuation time is calculated as > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 3 TABLE II ET = T – BT (1) ET: evacuation time T : ticks count BT: bomb time. The time before the bomb goes off. this case 10 We assumed this market is an open one. So we didn’t have any obstacles (other buildings) in this system. STATISTICAL DESCRIPTIVE OF OUTPUT VARIABLES Output Variables Standard Deviation Mean Minimum Maximum ND 8.1 2.8 2 17 NI 7.9 2.7 1 14 NR 22.6 6.8 1 41 ET 122.8 12.1 91 151 ND = number of deaths, NI = number of injuries, NR = number in rescue area, ET = evacuation time. IV. ANALYSIS AND DISCUSSION The purpose of this research is to provide a preliminary attempt at using agent-based modeling to investigate the impact of structural layout, people behavior, and environmental conditions on the evacuation of people from a marketplace. In order to investigate the effect of such elements on evacuation, a set of input variables were loaded to the NetLogo model (Table 1). The model was run a total of 300 simulation runs. In every run, both the location of shoppers and the rescue area have been changed. Fig. 2 illustrates a snapshot of a simulation run during the evacuation process. Table 2 summarizes the statistical descriptive of the output variables in relation with the location of the rescue area. Since the location (coordinates) of the rescue area has been randomly changed at every simulation run, it is essential to further investigate the impact of such locations on the output variables, namely evacuation time, number of deaths, number of injured, and number of shoppers who run to the rescue area. Figures 3-6 depict the surface plots for the rescue area locations and the output variables. TABLE I 15 MODEL INPUT VARIABLES Input Variables Values / Coordinates Population 150 Security officers 20 Bomb impact 12 Bomb location (x, y) (-19, 17) Bomb time 10 Check point location (x, y) (0, 35) Run to rescue area distance 14 ND = number of deaths, NI = number of injuries, NR = number in rescue area, ET = evacuation time. ND 10 5 0 -40 20 -20 0 0 X Y -20 20 -40 Fig. 3. Number of Deaths as a function of applied field. 15 NI 10 5 0 20 -40 0 -20 0 X 20 -20 -40 Fig. 4. Number of Injuries a function of applied field. Fig. 2. Snapshot of Netlogo after the bomb deployed. Y > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < 4 on people safety. Further, since any layout design (structure) TABLE III OUTPUT VARIABLES FOR DIFFERENT LOCATIONS Output Variables ND NI NR ET 45 30 NR 15 0 -40 L1 (-35, -21) 2* 12 14** 136 Output Variables -20 L2 (34, 30) 17** 9 19 132 L5 (2, -8) L3 (20, 10) 6 1* 36 119 L6 (28, -17) L7 (25, -17) L4 (-33, -19) 7 14 13 134 L8 (17, -18) 20 0 ND NI NR ET 0 X 20 Y -20 -40 Fig. 5. Number in Rescue Area. 9 5 1* 98 2* 10 41** 141 10 4 26 91* 15 13 31 151** ND = number of deaths, NI = number of injuries, NR = number in rescue area, ET = evacuation time. must provide protection to the people to reach safety, future research can be conducted to investigate the optimal location of such design components to minimize evacuation time and number of deaths and injuries. Finally, the deterioration of environmental conditions, level of uncertainty, stress, and the interaction of people with such factors influence the adoption of new responses. Such responses can be incorporated into advanced versions of the proposed agent-based model. 160 140 ET 120 100 -40 0 0 X ACKNOWLEDGMENT 20 -20 -20 20 Y -40 Fig. 6. Evacuation Time a function of applied field. The authors would like to thank Drs. Rani A. Kady ([email protected]) and Andreas Tolk ([email protected]), both with the department of Engineering Management and Systems Engineering, for their guidance and support throughout the study. REFERENCES V. CONCLUSION The purpose of this research was to provide a preliminary attempt at using agent-based modeling and simulation analysis to explore potential options that people have when faced with bomb explosion in an open market Environment. The developed model demonstrates the significant role agent-based modeling plays to capture accurate and representative details of the evacuation process at the agent (evacuee) level. Further, the results show that evacuation time, number of fatalities, and number of injuries are influenced by the layout design components, namely locations of rescue area, check point, and the bomb itself. Table 3 lists a range of statistical output measures for four different selected locations. [1] [2] [3] [4] [5] [6] VI. FUTURE RESEARCH The developed model can be used in future research to investigate allocating necessary resources such as ambulances, hospitals, shelters to minimize the impact of bomb explosions [7] ""Modeling and simulation methodologies: Real time simulation, System dynamics, Agent-based and modeling simulation." class notes for MSIM 601, Department of Modeling and Simulation, Old Dominion University, 2009. R. Lempert, "Agent-based modeling as organizational and public policy simulators," in Proc. of the national academy of sciences of the United States, vol. 99, pp. 7195–7196, May 2002. E. Bonabeau, "Agent-based modeling: Methods and techniques for simulating human systems," in Proc. of the national academy of sciences of the United States, vol. 99, pp. 7280–7287, May.2002. C.M. Henein, T. White &, "Agent-based modeling of forces in crowds," in Multi-Agent and Multi-Agent-Based Simulation, vol. 3415, P. Davidsson et al, Ed. Springer Berlin Heidelberg, 2005, pp. 173–184. A. Kirchner, A. Schadschneider, "Simulation of evacuation processes using a bionics inspired cellular automaton model for pedestrian dynamics. Physica A, vol. 312, 2002, pp 260-276. A. Shendarkar, K. Vasudevan, S.Lee, Y.J.Son, "Crowd simulation for emergency response using BDI agents based on immersive virtual reality," Simulation Modelling Practice and Theory, vol. 16, pp. 1415– 1429, 2008. X. Chen. (2003). Agent-based simulation of evacuation strategies under different road network structures [online]. Available: http://www.ucgis.org/summer03/studentpapers/xuweichen.pdf > REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < [8] C. Ren, C. Yang & S.Jin, "Agent-based modeling and simulation on emergency evacuation," in Complex Sciences, vol. 5, J. Zhou, Ed. Springer Berlin Heidelberg, 2009, pp. 1451–1461. [9] U. Wilensky, (1999). NetLogo. Center for Connected Learning and Computer-Based Modeling. Northwestern University, Evanston, IL [Online]. Available: http://ccl.northwestern.edu/netlogo/ [10] U. Wilensky. (1999). Northwestern University, Evanston, IL [Online]. Available: http://ccl.northwestern.edu/uri/ 5
© Copyright 2026 Paperzz