Building Egress Strategies

“A Critical Review of Emergency Evacuation Simulation Programs from a Social Science Perspective,” Gabriel Santos and Benigno Aguirre, University of Delaware

The paper presents a critical review of selected simulation models, including (1) flow based, (2) mesoscopic, (3) cellular automata, (4) agent-based, and (5) activity-based models, as well as of three simulation models--FIRESCAP, EXODUS, and the Multi-Agent Simulation for Crisis Management--that incorporate social scientific processes. It concludes by pointing out the so far ignored insights that could be derived from a dual emphasis on the social psychology of the actor and on macro social organizational features such as norms and values shown in Turner and Killian's (1987) emergent norm theory of collective behavior. It concludes with a number of predictions derived from ENT regarding the effects of social organizational variables on the timing of evacuation behavior.

• Presentation visuals
• A Critical Review of Emergency Evacuation Simulation Models, Gabriel Santos and Benigno Aguirre, University of Delaware
• Disaster Research Center at the University of Delaware

“WTC Evacuation Study,” Robyn Gershon, Columbia University, NY.

Data on the qualitative phase of the WTC Evacuation Study will be presented. Special emphasis will be placed on the individual, organizational and structural factors that affected evacuation. Preliminary findings based on the data will be discussed.

• Presentation visuals
• The World Trade Center Evacuation Study at Columbia University

“Public Behavior in Response to Warning Information,” Dennis Mileti, Natural Hazards Research Center, University of Colorado

This presentation will refer attendees to the over 300 publications in the social sciences regarding public response to risk information. It will summarize this literature regarding the state of the art for public warning response (which includes evacuation) and pre-emergency public education. Bibliographies and summary papers will be made available to workshop participants.

• Bibliography of non-high rise/disaster research “evacuation” literature
• Annotated bibliography of non-high rise/disaster research "evacuation" literature
• Pre-event Public Education: Summary of non-high rise/disaster literature (written in the form of a check list for people who would use the knowledge and based on the above referenced bibliography).
• General Evacuation Theory: Summary of non-high rise/disaster literature on public response to emergency warning information (written as a series of slides for a training course and based on the above referenced bibliography)
• Natural Hazards Center at the University of Colorado

“Developing Robust Evacuation Instructions within an Intelligent Evacuation, Rescue and Recovery System,” Elise Miller-Hooks, University of Maryland

A concept for an expert system will be described that, through the use of sensor technology, can permit real-time assessment of the extent of blast damage to a building, can recommend immediate actions that can be taken to mitigate the situation and prevent further deterioration, and can be used to aid the rescue workers and evacuees in rescue efforts and safe egress. The key capabilities of this system stem from the electronic integration of two critical components: a near real-time intelligent BDA/TVA tool and on-line egress-related optimization techniques. Methodologies will be discussed for determining optimal and robust tactical and operational strategies for rapidly evacuating a large burning building or a building that has come under attack by enemy or natural catastrophe. These procedures explicitly consider the inherent dynamic and uncertain nature of circumstances requiring evacuation. Therefore, they give rise to robust evacuation plans with lower probability of failure than paths determined otherwise, enabling faster and more efficient evacuation of a building in the event of military attack, fire, natural disaster, discovery of a hazardous material or biological agent, or other circumstances warranting quick escape.

• Presentation visuals

The Challenge of Creating Protective Environments: Leveraging Information for Dynamic Feedback, Gregory Luther, United Technologies Research Center

Building systems are increasingly equipped with sensors, communications and computing technologies to deliver cost effective operations, security, and building management services. Advancements in building state awareness through wireless sensor networks, video surveillance, and RF ID’s can be leveraged using data mining and reasoning algorithms that enable emergency personnel to manage large amounts of data more quickly in order to make higher fidelity decisions that will protect people from harm. This sequence of awareness, communication, planning and action comprises a feedback loop that creates and maintains dynamic protective environments. It must be considered in the design and implementation of effective egress strategies. Modeling capabilities for threat evolution and occupant motion, an integrated concurrent design process, and demonstration projects are necessary elements for progress towards safer building systems.

• Presentation visuals

“Sensor Networks and Elevator Control for Optimizing Building Evacuation,” Christos Cassandras, Boston University

Emerging wireless sensor network technologies provide unprecedented opportunities for "smart" building management leading to lower costs, higher efficiency, and better security. In emergency situations, sensor networks can play a critical role in supplying real-time information which can be used in conjunction with appropriate elevator dispatching control for building evacuation. This presentation will overview modeling and analysis methodologies required to actualize the potential of new technologies for smart building management, discuss the problem of optimal evacuation, and show how its solution is sometimes counterintuitive.

• Presentation visuals

“Coherent Configuration and Operation of Building Traffic Systems,” Shi-Chung Chang (presenting), Peter Luh, Bo Xiong, and Laurent Michel, University of Connecticut

Life-cycle optimization of building traffic systems is to exploit various technologies for the configuration, operation, and adaptation of these systems for improved efficiency, reduced costs, enhanced security, and greater occupant satisfaction. Formal models, methods, and tools are largely missing even in their outline forms. Using normal and egress modes of elevators as a conveyer example, this paper presents a methodological framework for coherent configuration and operation optimization of building traffic systems. In the framework, formal semantics serves the coherence specification in a multi-model, multi-mode, and people-in-loop environment. A price-based decomposition and coordination approach then solves the problem for both normal and emergency modes while considering the interactions with HVAC systems.

• Presentation visuals

Session Summary

The objective of this session was to discuss the use of predictive models and emerging technologies in the development of new evacuation strategies. The emphasis was on tall buildings. Three main themes emerged from the presentations and discussions: a large body of knowledge in the area of human response in emergencies is available but largely untapped, the ability to develop predictive models is available but data is required for validation, new technologies are becoming mature to the point where they can be used to improve occupant safety and provide early information to first responders but technology development and demonstration projects need to be appropriately resourced.

The current state of knowledge in the area of human response related to egress conditions is represented by several studies conducted over the last decades on emergency evacuation of cities, mostly in the context of earthquakes and nuclear disasters. One part of these studies was focused on developing an understanding of human response and the factors that influence it. In recent years, studies have focused on building and ship egress, however the number of such studies is still small. Several speakers suggested that this body of knowledge remains largely untapped by the community studying building egress. In the current state of the practice evacuation plans are developed to be executed during emergencies. These plans do not rely on the use of real-time information or knowledge of how to influence human behavior in stressful situations to guide occupant egress.

Several models have been developed in recent years to predict human behavior during emergencies. There is a wide variation in what aspects of human behavior are accounted for, ranging from pure kinematics, to flow-type models, to cellular-automata, to rule-based models. One common thread in the discussions is the unavailability of reliable data to independently validate existing and future models. Another aspect is the ambiguity regarding the required level of accuracy as a function of the questions being investigated. For example, is a flow model sufficient for the design of egress routes or should more complex social interactions such as leader behavior be included? For a given application, how does one establish that models provide the required accuracy level? Lastly, when should models be used, in design of high-profile buildings or in more routine building and egress design?

Several speakers addressed emerging technologies and their potential use as a means to achieve safe and controlled building egress. Projects sponsored by the Department of Defense over the past two decades have produced technologies that could be adapted to increase building safety. Examples are sensor networks, large-scale optimization methods, distributed control and actuation, and supervisory control. These and other technologies have been successfully used to enable remote sensing of battlefields, perform real-time threat assessment, and control autonomous flying drones. However, most of these technologies are too expensive or require adaptation for deployment in typical buildings.

The discussion highlighted the need to address the main obstacles first while producing results that could be used by first-responders in the short-term. The main obstacles identified were the following:

• The gap between social sciences and engineering needs to be bridged. The effectiveness of new technologies is dependent on their ability to be integrated in inhabited environments and used by building occupants. For example, how do we use existing knowledge and technologies to direct occupant behavior in a desirable manner? What procedures can rely on the automated operation of building systems and which cannot?
• Reliable and readily available experimental data needs to be measured and used to validate predictive models. Technological and social barriers to acquire the data were mentioned, such as occupants’ reluctance to participate in drills and the absence of automated ways to record traffic flow that extract the necessary information for model development.
• There is a lack of models capable of accurately predicting the population response during an evacuation. For example, there are several studies on the possible use of elevators during egress situations with widely different conclusions on its feasibility. Discrepancies like this decrease the confidence in the use of models as tools to develop new egress strategies.
• Full-scale test sites are needed to demonstrate solutions. The adoption of new solutions would be faster if demonstration projects in inhabited buildings were pursued and used to gather feedback from stakeholders.

The participants in this session made several core recommendations. They identified a basic need to create mechanisms that enable joint work on building egress among national laboratories, academic researchers and industrial concerns. They identified the need for key stakeholders to develop a roadmap that articulates a development and demonstration path for new solutions. Both funding and test sites would be identified as part of the execution of this vision. Finally, they identified the need to tackle several urgent projects of smaller magnitude in the near term. These projects will address central parts of a longer-term vision, but they have high urgency and longer lead-times. These near-term initiatives and their stakeholders should be identified along with the appropriate funding mechanisms that will bring teams and resources together.

Privacy Statement/Security Notice  |  Disclaimer  |  FOIA NIST is an agency of the U.S. Commerce Department's Technology Administration.