Evaluating and Improving the Safety of Building€¦ · Psychology of the Event •From a game theory perspective - staying calm is not a stable strategy. •Egress under panic is

Evaluating and Improving the Safety of Building Layouts for Active Assailant ThreatsPriority Subject Matter: Architecture, engineering & design

IntroductionEvaluating and Improving the Safety of Building Layouts for Active Assailant Threats

Tim Brewer, CEng MICE

• Professional structural engineer

• Specializing in physical protective security design

• Focus on terrorism: structural & human vulnerability

• Government and commercial projects

• Industries: education, sport, energy, transportation

Abstract

• Hypothesis: the application of risk-based analytical approaches can improve the way in which we evaluate alternative security designs.

• Approach: Leveraging occupant egress simulations to compare and contrast different building layout options for credible active assailant threat scenarios.

Learning Outcomes

1. Discover potential shortcomings of purely experience-based design decisions

2. Understand egress analysis and its application to active assailant scenario problems

3. Recognize optimum design layouts and identify other ways to improve safety such as areas of refuge and mass notification systems

Business Impact

• Improve the means and methods for security practitioner evaluation of building layouts for armed assailant threat scenarios.

• Panicking people will not listen to a person yelling: “everybody stay calm!”, so a solution may be to redesign the architecture to maximize building occupant egress.

• Ideally, this improvement would be incorporated from the initial design of a building, but the exercise is also useful during remodeling and even assessment work (i.e., maybe occupancy should be reduced, or changes made to the current layout)

• I recognize that architects/designers are often best placed to effectuate these changes, but as Security Professionals, you are best placed to provide, advice, guidance, and SME input.

…..some clarifications

• I am going to discuss the movement of individuals and groups in a state of panic and the implications of the physical layout of buildings and floor plans on this movement in the context of an armed aggressor / active assailant scenario.

• Whilst acknowledging the relevance to architects and engineers, much of the material will be useful and interesting to security practitioners. You will be viewing this subject from a new perspective.

• Although the narrative may often consider schools, this briefing is meant to address a broad spectrum of building use types. Often these principles are best applied in high-density occupied spaces and crowds (e.g., sporting stadia and transportation hubs)

Alert System

Access Control

Intrusion Detection

Video Surveillance

Hardened Shelters

Gunshot Detection

Training & Drills

Mass Notification

Building Layout

Active Shooter / Armed Assailant:A complex problem requires an integrated security solution:

Engineering

Psychology

Security

Some reasons not to leave!

• Not a software or consultancy sales pitch

• Some videos later

• Real-life quantitative case-studies and examples

• Excellent references and data sources

• Please feel free to ask questions throughout and make this interactive

Format

• Data & Statistics– what’s it telling us?

• Current & Future Guidance

• Event – what happens?

• Leveraging Modeling to Support Security Design

• Lessons for Security Design

• Conclusions

Data & Statistics – what’s it telling us?Evaluating and Improving the Safety of Building Layouts for Active Assailant Threats

TerminologyTerminology Agency Definition

Mass ShootingCongressional Research

Service (CRS)More than four people are killed with a firearm within one event, and in one or more locations in close proximity

Mass Killing Congress Three or more killings in a single incident

Active Shooter FBIAn individual actively engaged in killing or attempting to kill people in a populated area

Armed Assailant

Armed Aggressor

How many ?

0

5

10

15

20

25

30

2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020

Nu

mb

er o

f A

ctiv

e Sh

oo

ter

Even

ts*

*As defined by the U.S. Department of Justice, Federal Bureau of Investigation

Where?

Businesses73, 46%

Schools39, 24%

Government properties16, 10%

Open spaces15, 9%

Residences, 7, 4%

Houses of worship, 6, 4%

Health care facilities, 4, 3%

Source: FBI, A Study of Active Shooter Incidents in the United States Between 2016 and 2017

57% of attackers are known to the victims

74% of attackers walked through the main entrance

37% of the attacks were over in under 5 minutes

63% of the attacks were over in under 15 minutes

5 10 150

• National Average Police Response Time 7-15 Min

• So, we need to facilitate survival or escape

(Nobody has been shot behind a locked classroom door)

What happens?Source: FBI, A Study of Active Shooter Incidents in the United States Between 2016 and 2017

Current & Future GuidanceEvaluating and Improving the Safety of Building Layouts for Active Assailant Threats

• NFPA released the world’s first active shooter/hostile event standard with guidance on whole community planning, response, and recovery

• Developed with insight from law enforcement, fire, EMS, medical providers, facility managers, private industry, DHS, the CIA, FBI and others

• The scope of this standard is limited to the necessary functions and actions related to preparedness, response, and recovery from an active shooter/hostile event response (ASHER). This standard applies to any community, authority having jurisdiction (AHJ), facility, and member of any organization who responds to or prepares for ASHER incidents.

• Excellent procedural advice

• Limited information regarding physical security

ASTM Standard for Armed Aggressors (Mass Violence) in Educational Institutions

• Risk Assessment

• Perimeters

• Layouts

• Building Components

• Operations

• Drills & Exercises

• Coming in 2019

Event – what happens?Evaluating and Improving the Safety of Building Layouts for Active Assailant Threats

Basically…Security 101

Threat

Get out!

Shelter (protect)

Decision makingOn the evening of 1 March 2014, a terroristattack occurred inside the Kunming RailwayStation in Kunming, Yunnan, China. Ataround 9:20 p.m., a group of eight knife-wielding men and women attackedpassengers at the city's railway station.

Both male and female attackers pulled outlong-bladed knives and stabbed and slashedpassengers. At the scene, police killed fourassailants and captured one injured female.

The incident left 31 civilians and 4perpetrators dead with more than 140others injured.

• Attackers (circled)

• General Panic

• Watch him

Individuals tend to develop blind actionism (unevaluated activity)

People move considerably faster than normal

Individuals start pushing, and interactions among people become physical in nature

Moving and in particular passing through a bottleneck becomes

uncoordinated

At doors and exits, clogging and arching are observed

Physical interactions in crowds can cause pressures up to 650psi,

enough to bend steel

The magnitude and direction of the forces can change rapidly,

pushing people in an uncontrollable way

Escape is slowed down by fallen or injured people turning into

”obstacles”

People show herding behavior, i.e., interacting agents transfer

contagious effects

Alternative exits are not efficiently used or overlooked

during escape situations

People try to escape from the source of panic, which can be

dynamic and often not related to the threat

Event – what happens?

Psychology of the Event

• From a ‘game theory’ perspective - staying calm is not a stable strategy.

• Egress under panic is often compared to the ‘prisoner’s dilemma’ that shows why two completely rational individuals might not cooperate, even if it appears that it is in their best interests to do so

• ‘common goods dilemma’ The dilemma arises when members of a group share a common good (e.g., the ability to exit a building). Anyone can use the resource but there is a finite amount of the resource available and it is therefore prone to overexploitation, where in a non-iterated game, it is reasonable for the players to take selfish actions rather than cooperate

Development of crowd behavior theories

1895

• The Irrationalist Approach (Le Bon, 1895)• The “Panic” Model

1960

• Emergent Norm Theory (Turner R., 1974)• More rationalist approach

Present

• The Social Identity Model (Reicher, 2001)• In emergencies, people seek out attachment figures: social norms rarely break down

• But, such ties can have fatal consequences- people escape (or die) in groups

Some crowd modeling myths

• Myth of Irrationality : the idea that individuals in a crowd lose rational thought

• Myth of Emotionality : the idea that individuals in a crowd become more emotional

• Myth of Suggestibility : the idea that individuals in a crowd are more likely to obey or imitate

• Myth of Destructiveness : the idea that individuals in a crowd are more likely to act violently

• Myth of Spontaneity : the idea that in a crowd violence occurs more suddenly

• Myth of Anonymity : the idea that individuals in a crowd feel more anonymous

• Myth of Uniformity/Unanimity : the idea that all individuals in a crowd act in the same way

Path Cost Calculation

Destination Selection

Path Choice Stage

Decision-making and behavior model for the onset of an attack

Initial Judgement

Evacuate Decision

Threat Information

NoPath Choice

Yes

Initial Judgement Stage

Assessment of Threat Strength

Layout of the Area

Global Assessment Stage

S. Li, J. Zhuang, and S. Shen, “A three-stage evacuation decision-making and behavior model for the onset of an attack,” Transportation Research Part C: Emerging Technologies, vol. 79, pp. 119–135, Jun. 2017.

How can we incorporate all this during security design/assessment?

1. Understand how the environment (e.g., building) effects the movement of people in a state of panic – crowd modeling

2. Identify potential problem areas (e.g., choke points)

3. Provide mitigating measures1. Operational control

2. Engineering works

Leveraging Modeling to Support Security DesignEvaluating and Improving the Safety of Building Layouts for Active Assailant Threats

Can we model crowd movement effectively?

Inputs:

• Individual physical characteristics of crowd members (e.g., size)

• Behavior under conditions of panic (e.g., proximity comfort levels, force of movement, speed, route selection etc.)

Phenomenology:

• Crowd self organizing effects when building occupants at risk or under evacuation conditions:• Striping effect• Zipper effect• Herding effect• ‘Faster-is-slower’ effect

IN

EXIT

IN

EXIT

IN

EXIT

IN

EXIT

Lessons for Security DesignEvaluating and Improving the Safety of Building Layouts for Active Assailant Threats

EXIT

EXIT

10% faster evacuation time, 30% lower risk of injuriesExample 1: Optimal door location, mid-wall or corner?

EXIT

EXIT

7% faster evacuation time, 70% lower risk of injuriesExample 2: Filtering methods for egress, safer/faster?

Example 3: Classroom/office desks – can we optimize egress with desk layouts?

-20

-15

-10

-5

0

5

% In

crea

se /

Dec

reas

e in

Eva

cuat

ion

Tim

e

Exam

ple

4: D

oo

r av

aila

bili

ty a

nd

loca

tio

n

Example 5: Office floor plan – availability of exits, proximity to attackers

41

Example 6: Modeling campus style events

Conclusions

1. Identified potential shortcomings of purely experience-based design decisions

2. Understand egress modeling and application to active assailant scenario problems

3. Recognize some optimum design layouts

Architects/designers are often best placed to building changes, but as Security Professionals, you are best placed to provide, advice, guidance, and SME input.

Hypothesis: the application of risk-based analytical approaches can improve the way in which we evaluate alternative security designs.

References

• A. Cuesta, E. Ronchi, S. M. V. Gwynne, M. J. Kinsey, A. L. E. Hunt, and D. Alvear, “School egress data: comparing the configuration and validation of five egress modelling tools: School Egress Data,” Fire and Materials, vol. 41, no. 5, pp. 535–554, Aug. 2017.

• S. Chen, Y. Di, S. Liu, and B. Wang, “Modelling and Analysis on Emergency Evacuation from Metro Stations,” Mathematical Problems in Engineering, vol. 2017, pp. 1–11, 2017.

Evaluating and Improving the Safety of Building Layouts for Active Assailant ThreatsPriority Subject Matter: Architecture, engineering & design

Armed Aggressor Threat within a Terminal BuildingModeling of a potential armed aggressor scenario facilitates the

calculation of egress times, and occupant risk, as well as identifying

potential choke points and supporting emergency response planning.