Lessons from the evacuation of the world trade centre, 9/11 2001 for the development of computer-based simulations

This paper reviews the state-of-the-art in evacuation simulations. These interactive computer based tools have been developed to help the owners and designers of large public buildings to assess the risks that occupants might face during emergency egress. The development of the Glasgow Evacuation Simulator is used to illustrate the existing generation of tools. This system uses Monte Carlo techniques to control individual and group movements during an evacuation. The end-user can interactively open and block emergency exits at any point. It is also possible to alter the priorities that individuals associate with particular exit routes. A final benefit is that the tool can derive evacuation simulations directly from existing architects, models; this reduces the cost of simulations and creates a more prominent role for these tools in the iterative development of large-scale public buildings. Empirical studies have been used to validate the GES system as a tool to support evacuation training. The development of these tools has been informed by numerous human factors studies and by recent accident investigations. For example the 2003 fire in the Station nightclub in Rhode Island illustrated the way in which most building occupants retrace their steps to an entrance even when there are alternate fire exits. The second half of the paper uses this introduction to criticise the existing state-of-the-art in evacuation simulations. These criticisms are based on a detailed study of the recent findings from the 9/11 Commission (2004). Ten different lessons are identified. Some relate to the need to better understand the role of building management and security systems in controlling egress from public buildings. Others relate to the human factors involved in coordinating distributed groups of emergency personnel who may be physically exhausted by the demands of an evacuation. Arguably, the most important findings centre on the need to model the ingress and egress of emergency personnel from these structures. The previous focus of nearly all-existing simulation tools has been on the evacuation of building occupants rather than on the safety of first responders. Thanks are due to J. Appleby, P. Cooper, A. Foss, S. Hailey and B. Jenks who were responsible for the design and implementation of the GES application. They also drove the development of the Boyd Orr evacuation scenarios that are used to illustrate the opening sections of this paper.     

ViCTS: A novel network partition algorithm for scalable agent-based modeling of mass evacuation

Emergency evacuation is a critical response to deadly disasters such as hurricanes, floods, and earthquakes, etc. However, mass emergency evacuation itself is a complex process that sometimes could lead to chaotic situations and unintended consequences. In many emergency scenarios, mass evacuation is necessary to cope with severe public threats within tight spatiotemporal ranges. To better understand complex phenomena like mass evacuation, and study possible consequences, agent-based models (ABMs) have been widely developed in previous work. Existing models simulate individual behaviors, posing computational challenges when applied to large geographic areas and sophisticated behaviors. A key strategy for resolving such computational challenges is to partition transportation networks into smaller regions and resolve corresponding computational costs by taking advantage of advanced cyberinfrastructure and cyberGIS. In this study, a novel network partition algorithm is developed to improve the scalability of agent-based modeling of mass evacuation based on a cutting-edge cyberGIS-enabled computational framework that exploits the spatial movement patterns of emergency evacuation. Specifically, the algorithm is termed as Voronoi Clustering based on Target-Shift, or ViCTS. It is enlightened by network Voronoi diagrams and designed to resolve computational scalability challenges caused by the unique characteristics of evacuation traffic. We conducted a set of computational experiments with real street network data in various evacuation scenarios to test the effectiveness and efficiency of the algorithm. Computational experiments show that ViCTS outperforms a widely used network partition algorithm for microscopic traffic simulation in terms of achieving optimal computational performance by balancing computational loads and reducing communications across high-performance parallel computing resources.     

Investigation into issues of passenger egress in Ladbroke Grove rail disaster

This paper provides an overview of findings from the Health and Safety Laboratory investigation into human factors issues surrounding passenger escape and evacuation in the Ladbroke Grove train crash (1999). Drawing upon a range of contextual information the investigation focused upon identifying barriers and delays to evacuation following the incident. Findings highlight a number of important design shortcomings, as well as scope for the enhancement of both passive and active communication systems. The need to take account of passengers’ mental models, intuitions and situation awareness when designing communication systems and emergency egress equipment for the railway carriage environment is highlighted.     

SENSITIZING SOCIAL AGENTS FOR VIRTUAL TRAINING

ABSTRACT

Virtual training allows the learning and rehearsal of implicit cues, e.g., trustworthy leading action in an emergency evacuation, that cannot be easily understood through merely reading about situations, while mitigating the danger and expense of live rehearsals. We have focused our efforts on designing social agents that can engage in and help to train humans to generate the trustworthy behaviors that help to ensure a successful evacuation. Drawing upon social science research and using a “role-reversal method,” we successfully constructed agents that can perceive trustworthiness as humans do. The agents first collect human responses to their own nonverbal cues in controlled experimental training scenarios. Using these results, we obtain optimal parameters for nonverbal cues of trustworthiness, and then can use them to guide agents who evaluate human performance in the same training scenarios. The method enables us to convert social psychological findings into computational mechanisms.     

基于ACP方法的应急疏散系统研究

目前地铁、建筑及煤矿等人员密集空间突发事件频发,解决人员密集空间人员安全疏散是当前社会发展亟待解决的问题. 本文以地铁为例研究了基于ACP方法的平行应急疏散系统,该系统由应急疏散人工系统、计算实验 和平行执行三部分组成. 平行应急疏散系统不仅适用于地铁,还适用于建筑及煤矿等生活和工作空间应急疏散管理,可实现人员培训、疏散演练、疏散方案优化和评估等功能. 本文初步建立了地铁枢纽站平行应急疏散系统,并进行了两种典型场景的计算实验,实验结果表明平行应急疏散系统可有效提高应急疏散效率.     

Modeling human behavior during emergency evacuation using intelligent agents: A multi-agent simulation approach

It is costly and takes a lot of time for disaster employees to execute several evacuation drills for a building. One cannot glean information to advance the plan and blueprint of forthcoming buildings without executing many drills. We have developed a multi-agent system simulation application to aid in running several evacuation drills and theoretical situations. This paper combines the genetic algorithm (GA) with neural networks (NNs) and fuzzy logic (FL) to explore how intelligent agents can learn and adapt their behavior during an evacuation. The adaptive behavior focuses on the specific agents changing their behavior in the environment. The shared behavior of the agent places an emphasis on the crowd-modeling and emergency behavior in the multi-agent system. This paper provides a fuzzy individual model being developed for realistic modeling of human emotional behavior under normal and emergency conditions. It explores the impact of perception and emotions on the human behavior. We have established a novel intelligent agent with characteristics such as independence, collective ability, cooperativeness, and learning, which describes its final behavior. The contributions of this paper lie in our approach of utilizing a GA, NNs, and FL to model learning and adaptive behavior of agents in a multi-agent system. The planned application will help in executing numerous evacuation drills for what-if scenarios for social and cultural issues such as evacuation by integrating agent characteristics. This paper also compares our proposed multi-agent system with existing commercial evacuation tools as well as real-time evacuation drills for accuracy, building traffic characteristics, and the cumulative number of people exiting during evacuation. Our results show that the inclusion of GA, NNs, and fuzzy attributes made the evacuation time of the agents closer to the real-time evacuation drills.     

Agent-based simulation of affordance-based human behaviors in emergency evacuation

Complex cognitive processes corresponding to human control behaviors cannot be easily inferred using (1) a logical rule-based model, (2) a statistical model, or (3) an analytical predictive model. Predicting human behaviors in complex and uncertain environments like emergency evacuation is considered almost impossible (at least NP hard) in systems theory. In this paper, we explore simulating human behaviors using affordance-based finite state automata (FSA) modeling, based on the ecological concept of affordance theory. To this end, we introduce the conceptual and generic framework of affordance-based human behavior simulation developed through our previous work. Following the generic framework, formal simulation models of affordance-based human behaviors are developed, especially for emergency evacuation, to mimic perception-based dynamic human actions interacting with emergent environmental changes, such as fire. A “warehouse fire evacuation” case is used to demonstrate the applicability of the proposed framework. The human action planning algorithms in the simulation model are developed and implemented using the Adjusted Floor Field Indicators, which represent not only the evacuee’s prior knowledge of the floor layout but the perceivable information about dynamic environmental changes. The results of our simulation study verify that the proposed framework accurately simulates human fire evacuation behavior. The proposed framework is expected to capture the natural manner in which humans behave in emergency evacuation and enhance the simulation fidelity of analyses and predictions of perceptual human behaviors/responses in the systems by incorporating cognitive intent into human behavior simulations.     

Scalable context-dependent analysis of emergency egress models

Pervasive environments offer an increasing number of services to a large number of people moving within these environments, including timely information about where to go and when, and contextual information about the surrounding environment. This information may be conveyed to people through public displays or direct to a person’s mobile phone. People using these services interact with the system but they are also meeting other people and performing other activities as relevant opportunities arise. The design of such systems and the analysis of collective dynamic behaviour of people within them is a challenging problem. We present results on a novel usage of a scalable analysis technique in this context. We show the validity of an approach based on stochastic process-algebraic models by focussing on a representative example, i.e. emergency egress. The chosen case study has the advantage that detailed data is available from studies employing alternative analysis methods, making cross-methodology comparison possible. We also illustrate how realistic, context-dependent human behaviour, often observed in emergency egress, can naturally be embedded in the models, and how the effect of such behaviour on evacuation can be analysed in an efficient and scalable way. The proposed approach encompasses both the agent modelling viewpoint, as system behaviour emerges from specific (discrete) agent interaction, and the population viewpoint, when classes of homogeneous individuals are considered for a (continuous) approximation of overall system behaviour.     

人群应急疏散可视仿真研究进展和问题

人群应急疏散可视仿真是用智能体来模拟具有自主感知、情绪和行为能力的人群个体,并采用3维可视的方式来直观呈现人群应急疏散情景,可以为制定人群应急预案提供形象直观的分析方法。本文从人群仿真数据的来源、人群导航模型的构建、人群行为模型、人群情绪感染、人群渲染5个方面概述目前研究的进展,然后从仿真模型的可验证性、人群疏散导航模型的构建、人与环境的物理模型、动物逃生实验与仿真、疏散中的社会行为表现以及人群情绪的可视计算6个角度讨论需要进一步研究的问题。针对需要深入研究的问题,指出借助于紧急事件的视频监控分析和虚拟人群情景的用户调查,有助于完善人群仿真模型。结合物理模型,可以更准确地描述人群应急疏散场景。开展动物逃生实验分析,有助于完善人群运动导航算法。建立人群社会行为模型,可以更详细描述疏散中人群行为的多样性。构建基于多通道感知的人群情绪感染计算方法,可以详尽描述情绪感染的过程。人群应急疏散行为的可视仿真研究在城市的安全管理方面具有重要的应用前景,但其研究仍存在很多亟待解决的问题,综合地运用多学科知识,完善实验手段是进一步推动研究的关键所在。     

Incorporation of BIM-based probabilistic non-structural damage assessment into agent-based post-earthquake evacuation simulation

Earthquakes can cause severe damage to structural and non-structural elements of buildings; consequently, they pose high risks to human lives. To mitigate such risks, attention has been paid to enhancing the indoor environment for increased building safety. Yet little effort has been made to assess a building occupants' evacuation behaviors in response to damage to the indoor environment. This paper addresses this issue with a novel simulation framework that couples human behaviors with changes to the indoor building environment during post-earthquake evacuation. In particular, we present a building information modelling (BIM)-based prototype that simulates seismic damage to the non-structural indoor elements and visualizes its impacts on evacuation using a color-coded heat map. The simulated damage is then used as input to an agent-based model for post-earthquake evacuation. Using a probabilistic method to assess the non-structural elements' damage states, we are able to evaluate the impact of indoor damage on the evacuation process. We performed a trial of our prototype for a hypothetical earthquake in an educational building. The results revealed how the average evacuation time would increase as the earthquake intensity increases (from 38.6 s for the no-damage scenario to 122.9 for the highest-damage scenario). The proposed prototype has the potential to be joined with other tools, such as finite-element-based simulation, to incorporate structural analysis as well. Planners and designers can explicitly use our model's output to analyze the post-earthquake evacuation with the indoor non-structural damage to assess different building design geometries that increase the chances of a suitable evacuation process.     

一个面向紧急疏散的大规模人群运动模拟系统

对公共场所大规模虚拟人群疏散过程的模拟研究在理论和现实方面都具有重大的意义.本文介绍了一个面向紧急疏散的大规模人群运动模拟系统,包括系统框架、功能、规模等.其中详细讲解了系统的几个功能模块及其实现技术,主要有人群建模和仿真技术、场景建模技术、灾害现象模拟技术、人群疏散控制策略以及海量数据实时渲染技术.最后对人群疏散系统在未来需要解决的问题做了简单讨论.     

SEEVis: A Smart Emergency Evacuation Plan Visualization System with Data-Driven Shot Designs

Despite the significance of tracking human mobility dynamics in a large-scale earthquake evacuation for an effective first response and disaster relief, the general understanding of evacuation behaviors remains limited. Numerous individual movement trajectories, disaster damages of civil engineering, associated heterogeneous data attributes, as well as complex urban environment all obscure disaster evacuation analysis. Although visualization methods have demonstrated promising performance in emergency evacuation analysis, they cannot effectively identify and deliver the major features like speed or density, as well as the resulting evacuation events like congestion or turn-back. In this study, we propose a shot design approach to generate customized and narrative animations to track different evacuation features with different exploration purposes of users. Particularly, an intuitive scene feature graph that identifies the most dominating evacuation events is first constructed based on user-specific regions or their tracking purposes on a certain feature. An optimal camera route, i.e., a storyboard is then calculated based on the previous user-specific regions or features. For different evacuation events along this route, we employ the corresponding shot design to reveal the underlying feature evolution and its correlation with the environment. Several case studies confirm the efficacy of our system. The feedback from experts and users with different backgrounds suggests that our approach indeed helps them better embrace a comprehensive understanding of the earthquake evacuation.     

Investigating the influence of route turning angle on compliance behaviors and evacuation performance in a virtual-reality-based experiment

Route turning is one of the most essential and ubiquitous physical features in the complex building environment. Under the influence of route turning, evacuees’ approaching perspective to an emergency sign could vary, affecting their information perception and behavioral compliance during the evacuation. Although conventional simulation methods assess the effectiveness of the emergency sign in the visible region, they fail to consider evacuees’ wayfinding behaviors and interaction with the emergency sign. It remains unclear whether the route turning angle affects evacuees’ compliance for detecting and responding to the emergency sign. To investigate such an influence, a virtual-reality-based method for assessing human evacuation behaviors in building fire evacuations was proposed. In this study, two evacuation routes with different turning angles in a shopping mall were created and implemented in a virtual-reality environment, and 67 subjects participated in the immersive virtual-reality-based experiment. All participants took the two routes to find the nearest exit for evacuation in a fire event, aiming to evaluate the effect of the route turning angle on the evacuation process. The participants were asked to complete a questionnaire at the end of the experiment. Next, statistical analyses were conducted on evacuation results, information perception, and evacuation performance of the participants. The results indicated the route turning angle significantly affected participants' behavioral compliance with emergency signs. The results also suggested the route turning angle was influential on participants’ information perception and evacuation performance. Besides, a significant effect on rotation change, wayfinding pause, and speed deviation were observed. This study validates the effectiveness of investigating evacuees’ interaction with emergency signs using virtual-reality technology and has potential implications for complex building path planning and evacuation simulation modeling.     

See how they run: modeling evacuations in VR

A virtual reality application called Vegas lets users experiment in real time with evacuation scenarios. Colt Virtual Reality, Ltd., based in Hampshire, United Kingdom, developed Vegas (virtual egress analysis and simulation), which runs on a 486-based PC. Its designers say that other evacuation models lack the user-friendly visualization of Vegas, and they hope its graphic capabilities will make the data available and comprehensible to a wider audience. In particular, they hope it will allow people to see how quickly fire and smoke can spread, and how important fast and calm action can be in an emergency. Colt has used Vegas to predict egress times from an underground subway station and evacuation times from a cross-channel ferry, as well as from several other, more traditional, buildings     

Modeling cooperative and competitive behaviors in emergency evacuation: A game-theoretical approach

A game-theoretical model to study evacuees’ cooperative and competitive behaviors during an emergency evacuation is proposed. The model integrated with evacuation dynamics model determines the density of cooperative and competitive evacuees and their related evacuation times. Computer simulation results show that (1) as urgency of evacuation increases, cooperation among evacuees’ decreases; (2) in an emergency situation, individual hyper-rationality among evacuees diminishes evacuation efficiency; (3) the imitation effect enhances cooperation among evacuees, yet reduces evacuation efficiency. This study provides a methodological pattern to research crowd behaviors in emergency evacuation.     

基于ACP方法的高层建筑火灾中人员疏散策略研究

高层建筑结构规模巨大、人员众多,而疏散出口数量有限,如何有效安排疏散出口人员分布,提高疏散效率,是火灾应急管理的重要研究内容. 以人为中心的疏散系统是典型的复杂系统,具有难以真实实验分析的困难. 本文基于ACP（人工系统、计算实验、平行执行）方法,以智能体技术为核心建立了人工疏散系统,基于火灾场景,利用计算实验对疏散策略进行了验证、评估,给出了实际疏散系统与人工疏散系统的平行执行实现思想. 最后,通过案例验证了方法的可行性.     

Crowd simulation-based knowledge mining supporting building evacuation design

Assessing building evacuation performance designs in emergency situations requires complex scenarios which need to be prepared and analysed using crowd simulation tools, requiring significant manual input. With current procedures, every design iteration requires several simulation scenarios, leading to a complicated and time-consuming process. This study aims to investigate the level of integration between digital building models and crowd simulation, within the scope of design automation. A methodology is presented in which existing ontology tools facilitate knowledge representation and mining throughout the process. Several information models are used to integrate, automate and provide feedback to the design decision-making processes. The proposed concept thus reduces the effort required to create valid simulation scenarios by applying represented knowledge, and provides feedback based on results and design objectives. To apply and test the methodology a system was developed, which is introduced here. The context of building performance during evacuation scenarios is considered, but additional design perspectives can be included. The system development section expands on the essential theoretical concepts required and the case study section shows a practical implementation of the system.     

Insights toward robot-assisted evacuation

This paper considers the problem of human evacuation assistance. We discuss how this kind of task differs from the more prevalent search and rescue tasks, and the resulting implications for the design of assistive evacuation systems. We describe the implementation and evaluation of an algorithm for deploying audio navigational cues throughout an office building with a team of mobile robots. A review of evacuation dynamics methods is presented and particular methods are applied on-line during deployment. We use a pedestrian simulation and a simple model of audio–evacuee interaction to show the effects of beacon deployment; the results indicate that even a small number of beacons can significantly decrease the mean and variance of egress time and distance. We also advance uses for human motion-based measures of environment complexity for general mobile robotics.     

Optimization of Grouping Evacuation Strategy in High-rise Building Fires Based on Graph Theory and Computational Experiments

It is difficult to rescue people from outside, and emergency evacuation is still a main measure to decrease casualties in high-rise building fires. To improve evacuation efficiency, a valid and easily manipulated grouping evacuation strategy is proposed. Occupants escape in groups according to the shortest evacuation route is determined by graph theory. In order to evaluate and find the optimal grouping, computational experiments are performed to design and simulate the evacuation processes. A case study shown the application in detail and quantitative research conclusions is obtained. The thoughts and approaches of this study can be used to guide actual high-rise building evacuation processes in future.      

Modeling emergency evacuation of individuals with disabilities (exitus): An agent-based public decision support system

In this paper, we present a public decision support system (DSS) distinguished from various DSSs in the private business sector in terms of its ownership, data scarcity, and beneficiaries. In particular, our system is intended to play a vital role in assessing and optimizing emergency response plans for rare but catastrophic events such as the September 11th attacks. While taking an agent-based microscopic simulation approach in a hierarchical framework, we used our model to estimate the effectiveness of alternative evacuation strategies to support emergency response planning as a part of business continuity planning for all private business organizations. The presented model is unique because it considers individuals with disabilities explicitly in terms of speed, ability to negotiate the environment, and normalcy bias depending on type of disability. It is also capable of classifying the environment in terms of accessibility characteristics encompassing various conditions which have been shown to have a disproportionate effect upon the behavior of individuals with disabilities during an emergency.Through a series of simulation experiments, our system identified specific locations (e.g., the NW landing) on the 3rd floor of the test bed building as possible bottleneck spots under certain conditions (e.g., the sharp increase in individuals with disabilities among residents). This way, our system provides the architect with tools to test the structure’s design to determine how well it meets the identified requirements for emergency evacuation to accommodate this shifting demographic. In particular, our system strongly demonstrated the effectiveness of new emergency evacuation strategies for individuals with disabilities such as assisted evacuations which allows other healthy people to play more active roles compared to traditional strategies in which individuals with disabilities are helplessly waiting for assistants such as fire fighters at the designated area. Our system also revealed that people using wheelchairs and those with lower stamina were at the greatest risk. People with lower stamina such as the elderly, people with chronic health conditions, or those with temporary injuries are at a greater risk mainly because they are not easily identifiable. Ultimately, the proposed DSS system can be used to inform public policy professionals of more effective, evidence-based evacuation planning and environmental design methods based on a better understanding of the behavior of individuals with disabilities.