University of Southern California Dissertations and Theses

A multiattribute decision model for the selection of radioisotope and nuclear detection devices

(USC Thesis Other)

A multiattribute decision model for the selection of radioisotope and nuclear detection devices

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

A MULTIATTRIBUTE DECISION MODEL FOR THE SELECTION OF RADIOISOTOPE
AND NUCLEAR DETECTION DEVICES

by

Antonia Boadi

A Dissertation Presented to the
FACULTY OF THE USC SOL PRICE SCHOOL OF PUBLIC POLICY
UNIVERSITY OF SOUTHERN CALIFORNIA

In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF POLICY, PLANNING AND DEVELOPMENT

December 2021

Copyright 2021 Antonia Boadi

ii
Acknowledgments
I am amazed, humbled, and gratified on a daily basis by the benevolence of God,
particularly when it is manifested through the kindness of human beings.
I continue to be inspired by the life of my late friend, Art McGee. Art was a scholar,
activist, mentor, and overall mensch. I am a better person for having known him.
I have been blessed with committee members who embody brilliance, patience, and
kindness. Detlof von Winterfeldt, who conceived the idea for this study, has been a constant
source of inspiration; his insightful suggestions deftly guided the development of this study.
Richard John’s incisive comments provided me with a much deeper understanding of decision
theory. Azad Madni’s enthusiasm about the transdisciplinary nature of this work helped me to
recognize additional applications for this work in my professional life. Had it not been for my
committee, I might not have persevered during my illness.
Words cannot express my appreciation for Dean Juliet Musso, Dr. Debbie Natoli and Ms.
Anna Parks, who advocated for me when I lacked the physical strength and emotional fortitude
to do so for myself.
I am grateful to USC’s Center for Risk and Economic Analysis of Terrorism Events
(CREATE) for their financial and academic support during the initial stages of this work.

iii
Table of Contents
Acknowledgments ................................................................................................................................ ii
List of Tables ........................................................................................................................................ v
List of Figures ..................................................................................................................................... vii
Abstract ..............................................................................................................................................viii
Chapter 1: Introduction ........................................................................................................................ 1
Background .................................................................................................................. 1
Statement of the Problem and Organizational Context ............................................. 2
Purpose of the Dissertation ......................................................................................... 4
Methodology ................................................................................................................ 5
Research Questions...................................................................................................... 6
Intended Audience ....................................................................................................... 6
Use Case 1: Planning Activities ..................................................................... 7
Use Case 2: PRND Training........................................................................... 7
Use Case 3: Community Policing .................................................................. 8
Use Case 4: Federal Nuclear Incident Response ........................................... 9
Use Case 5: Military Response to International Crises ................................ 9
Use Case 6: EPA Radiological Response Team ......................................... 10
Assumptions and Limitations of the Study .............................................................. 10
Document Organization ............................................................................................ 11
Chapter 2: Literature Review ............................................................................................................. 13
Introduction ................................................................................................................ 13
Topic 1: History of the American Response to the Evolving Nuclear
Threat .......................................................................................................................... 13
Topic 2: Science and Technology............................................................................. 15
Nuclear and Radiological Weapons ............................................................. 15
Detection of Illicit Radiological and Nuclear Materials ............................. 17
Topic 3: Policy Constraints that Impact the Selection of Instrumentation ............ 18
Topic 4: Multiattribute Utility (MAU) Methodology ............................................. 22
Chapter 3: Methodology..................................................................................................................... 26
Development of MAU Model ................................................................................... 26
Definition of Objectives, Attributes, and Alternatives Research Methods ............ 27

iv
Objective 1: Maximize Detection Accuracy ............................................... 27
Objective 2: Maximize Operational Effectiveness ..................................... 28
Objective 3: Minimize Cost .......................................................................... 29
Evaluation of Alternatives Relative to Attributes .................................................... 29
Determination of Attribute Weights ......................................................................... 31
Swing Weight Procedure .............................................................................. 32
Evaluate Alternatives using Attributes and Weights ............................................... 33
Evaluate Alternatives using Attributes and Weights for non-Cost
Related Objectives ........................................................................................ 33
Evaluate Alternatives using Attributes and Weights for Cost
Attribute ......................................................................................................... 34
Summary .................................................................................................................... 38
Chapter 4: Evaluation of Backpack Radiation Detectors ................................................................. 39
Evaluation of Alternatives Relative to Attributes .................................................... 41
Calculation of Weights using Swing Weight Procedure ......................................... 42
Summary .................................................................................................................... 52
Chapter 5: Evaluation of Radioisotope Detection Device Portfolios .............................................. 54
Introduction ................................................................................................................ 54
Portfolio Optimization based upon Operational Scenario (Research
Question 2) ................................................................................................................. 55
Optimization .................................................................................................. 58
Summary .................................................................................................................... 64
Chapter 6: Conclusion ........................................................................................................................ 65
Assumptions and Limitations of Study .................................................................... 66
Future Work ............................................................................................................... 66
References ........................................................................................................................................... 68

v
List of Tables
Table 1. Body-Worn Devices ............................................................................................................. 21
Table 2. Human-Carried Devices ..................................................................................................... 22
Table 3. Summary of Attribute Measures ......................................................................................... 30
Table 4. Best–Worst Attribute Values ............................................................................................... 31
Table 5. Determination of Attribute Ranks and Weights ................................................................. 33
Table 6. Multiattribute Utility Model ............................................................................................... 34
Table 7. Utility vs. Price .................................................................................................................... 35
Table 8. Completed MAU .................................................................................................................. 38
Table 9. Consequence Table ............................................................................................................. 41
Table 10. Conversion of Consequences to Single Attribute Values ................................................ 42
Table 11. Calculating Weights for Non-Cost Related Attributes .................................................... 44
Table 12. Calculation of MAV for SPIRPack Device ...................................................................... 45
Table 13. Calculating Utility Values for Non-Cost Related Objectives ......................................... 46
Table 14. Unweighted vs. Weighted MAV (for Benefit Objectives) ................................................ 47
Table 15. Assignment of Values for BRD Price Using Swing Weighting Method ......................... 48
Table 16. Weights Assigned to Benefit and Cost Objectives ........................................................... 49
Table 17. Calculation of Overall MAV ............................................................................................. 50
Table 18. Overall MAV and Cost Data ............................................................................................ 51
Table 19. Device Catalog .................................................................................................................. 56
Table 20. Optimization Results for Law Enforcement Equipment Portfolio .................................. 58
Table 21. Score-Cost Ratios for Personal Radiation Detectors ..................................................... 59
Table 22. Score-Cost Ratios for Handheld Detection Devices ....................................................... 60
Table 23. Score-Cost Ratios for Backpack Detection Devices ....................................................... 60
Table 24. Device Portfolios for Funding Levels 25 - 45K .............................................................. 61

vi
Table 25. Device Portfolios for Funding Levels 50-75K................................................................. 62
Table 26. Device Portfolios for Funding Levels 80-100K .............................................................. 63
Table 27. Device Portfolios for Funding Levels 125-300K ............................................................ 64

vii
List of Figures
Figure 1. Organizational Chart of the Countering Weapons of Mass Destruction Office ............... 3
Figure 2. Hierarchy of Operational Attributes .................................................................................. 29
Figure 3. Utility vs. Cost .................................................................................................................... 35
Figure 4. Sensitivity Analysis (Utility vs. Cost Weight).................................................................. 36
Figure 5. Utility vs. Cost .................................................................................................................... 51
Figure 6. Sensitivity Analysis – Utility vs. Cost Weight ................................................................. 52
Figure 7. Law Enforcement PRND Team Requirements ................................................................. 57

viii
Abstract
The nuclear threat landscape has evolved from one dominated by the development and
deployment of conventional nuclear weapons to a more likely and easily executed scenario in
which an amorphous enemy launches an attack using a radiological dispersive device, more
commonly referred to as a dirty bomb. The U.S. has implemented a complex, multilevel nuclear-
detection architecture designed to deter, detect, and interdict attempts to transport contraband
material that can be used to construct nuclear/radiological weapons of mass destruction.
First responders support nuclear detection, interdiction, and crisis response missions. The
technologies that support these functions can be classified into five major archetypes: body-
worn, human-carried, portable, vehicle-mounted, and fixed. Collectively, these devices are
referred to as radioisotope detection devices. The selection of an appropriate device or portfolio
of devices involves multiple tradeoffs that are difficult for humans to perform without some kind
of decision aid. This recognition provided the motivation for this dissertation.
This dissertation provides decisionmakers with a tool that facilitates the selection of
radioisotope detection devices for use in nuclear/radiological crisis response missions. A
multiattribute utility (MAU) value model was used to identify criteria for determining the best
device or device portfolio for crisis response and routine monitoring. This model can be used to
guide planning, training and technology upgrade efforts implemented by local, tribal, state,
federal, and international responders to nuclear crises.

1
Chapter 1: Introduction
This dissertation is responsive to the need to evaluate the performance of radioisotope
detection devices in risk management scenarios. It is motivated by the recognition that selecting
a device or a portfolio of devices for nuclear detection, interdiction, and crisis response missions
involves multiple tradeoffs that are difficult for humans to perform without some kind of
decision aid. It presents a methodology for selecting devices for use in both the detection of
contraband materials as well for use in post-event scenarios.
Background
The American homeland security enterprise is a complex sociotechnical system. Both
qualitative and quantitative methodologies guide this investigation: (a) organizational theory,
including analysis of the concept of operations governing the responding organization, and
(b) decision theory, involving the development of mathematical models that support the
objective selection of alternatives based the operational scenario, timeline, event dynamics and
geographic considerations associated with the nuclear event.
The nature of the threat has evolved from the use of a nuclear weapon by a nation-state to
a scenario in which non-state actors or homegrown violent extremists (Department of Homeland
Security, 2019) execute an attack involving the use of a radiological dispersive device (RDD) or
improvised nuclear device (IND).
Several domestic and international programs have collaborated to protect the nation from
nuclear attack since World War II; they collectively evolved into the Global Nuclear Detection
Architecture (GNDA). The GNDA consists of the intelligence, equipment, processes, and
agreements that protect the United States from such attacks. The objectives of the GNDA are
(a) to protect existing stocks of nuclear materials and weapons from diversion or theft; (b) to

2
enhance domestic interdiction and detection efforts; (c) to eliminate excess stocks of nuclear
materials; and (d) to detect the illicit movement of radiological or nuclear materials overseas.
The Domestic Nuclear Detection Office (DNDO) was established in 2005 within the
Department of Homeland Security to oversee the domestic portion of the GNDA. DNDO was
tasked to develop a layered strategy for nuclear detection (Government Accountability Office,
2019). In 2017, DHS initiated a reorganization that consolidated chemical biological radiological
nuclear and high yield explosives (CBRNE) counter-terrorism functions with the Office of
Health Affairs. Although the GNDA is no longer associated with a particular directorate, the
GAO reports that the responsibility for analyzing performance gaps and vulnerabilities will be
distributed throughout the Countering Weapons of Mass Destruction organization (Government
Accountability Office, 2019).
Statement of the Problem and Organizational Context
Numerous devices for nuclear detection, interdiction, and crisis response are available on
the market. They differ in size, weight, capabilities, and cost, among other attributes. For those
with responsibility for acquiring and using these devices, it is extremely difficult to conduct a
comparative evaluation, considering the multiple conflicting attributes and complex tradeoffs.
This difficulty is exacerbated by the multiple purposes and needs for the use of these devices,
which may require purchasing a portfolio of devices to meet diverse needs.
The responsibility for developing, testing and evaluating devices for nuclear detection,
interdiction, and crisis response rested for many years with the Domestics Nuclear Detection
Office (DNDA) of the Department of Homeland Security. The Department of Homeland
Security’s Countering Weapons of Mass Destruction Act consolidated the Domestic Nuclear
Detection Office (DNDO) with organizations responsible for countering other WMD threats in

3
December 2018 (Countering Weapons of Mass Destruction Act, 2018). The reorganization
transferred the budget authority, personnel, and assets of DNDO and the Office of Health Affairs
to the Countering Weapons of Mass Destruction (CWMD) Office.
Figure 1.
Organizational Chart of the Countering Weapons of Mass Destruction Office

Note. Adapted from “Chemical Terrorism: A Strategy and Implementation Plan Would Help
DHS Better Manage Fragmented Chemical Defense Programs and Activities,” by Government
Accountability Office, 2018.

No specific directorate has been identified for coordinating the GNDA’s responsibilities
(Government Accountability Office, 2019; National Research Council, 2013). Because of the
reorganization, detection and interdiction activities may be performed in conjunction with crisis
response activities. First responders may need to perform radiation measurements and other
monitoring functions in the immediate aftermath of a nuclear or radiological attack, which could
take several forms, including the contamination of a food or water source, an attack on nuclear
power plants or nuclear waste storage facilities, the detonation of an improvised nuclear device,

4
or placing a radioactive source in a public venue (U. S. Department of Health and Human
Services, 2019).
The challenges associated with the initial stages of a radiological/nuclear crisis can
quickly overwhelm regional resources (National Council on Radiation Protection and
Measurement, 2011). Federal resources and responders will be deployed to perform extensive
assessment and monitoring 24 to 72 hours following a large-scale release (Musolino et al., 2019;
National Security Staff Interagency Policy Coordination Subcommittee for Preparedness &
Response to Radiological and Nuclear Events, 2010). In the event of simultaneous or cascading
incidents, federal responders could be significantly delayed. First responders may have to rely
upon radioisotope detection equipment should a nuclear or radiological crisis occur. Although
this equipment is in widespread use throughout the nation, it was designed to be sensitive to low-
levels of emissions while devices used in emergency response operations are designed to
generate alerts when environmental conditions exceed a predefined threshold. Conversely, health
and environmental monitoring functions aim to identify unacceptably high levels of radiation.
The dual nature of these performance objectives necessitates the development of a structured
approach for the selection of devices capable of supporting detection/interdiction as well as
emergency response missions.
Purpose of the Dissertation
This dissertation addresses the complexities in evaluating radioisotope detection devices
and in developing a cost-effective portfolio of instruments for multiple purposes and needs. It
proposes a methodology for selecting radioisotope detection devices for use in crisis response
missions. Crisis response activities support emergency relief and protect public health and safety
following a crisis involving a radiological/nuclear weapon of mass destruction (WMD). First

5
responders will have to use radioisotope detection equipment to support consequence
management functions until federal responders have been deployed to the local municipality,
typically within 24 to 48 hours (Buddemeier, Wood-Zika, et al., 2017; U. S. Department of
Health and Human Services, 2019). In the event of a catastrophic event, local responders may
have to rely upon radiation detection and dosimetry capabilities for extended periods.
It is important to note that this methodology does not recommend any particular device,
vendor, tool, or technology. Rather, it provides decision makers with a structured, defensible,
and transparent approach to acquisition decisions.
Methodology
Multiattribute utility (MAU) methodologies will be used to structure the complex
decision problem and then help decision makers select a radioisotope detection device that can
be deployed in interdiction, detection and consequence management missions (Edwards, 1971).
The output of the model is a ranked list of devices along with their relative scores.
The decision model consists of a set of candidate devices, the attributes that characterize
a device’s performance, operational, and cost performance, vendor device data, and a
weight/ranking associated with each attribute. The model uses the attributes described in
previous research (Buddemeir, Musolino, et al., 2016; Buddemeier, Wood-Zika, et al., 2017;
Musolino et al., 2019; U.S. Army, 2009). Subject matter experts (SME) assign weights based
upon their utility in both detection and crisis response missions. The procedure for eliciting the
opinions of the SMEs is described in Chapter 3. Chapter 4 presents a more thorough analysis on
a set of five candidate backpack radiation detectors. The expanded model encompasses both
PRND and crisis response scenarios. Chapter 5 applies the SMART methodology to the problem
of selecting an ensemble of detection devices.

6
Research Questions
This study investigates the issues associated with evaluating first responder preventive
radiological/nuclear detection (PRND) equipment for use in detection, interdiction, and disaster
response scenarios. PRND teams rely primarily upon three broad categories or archetypes of
detection instrumentation: body-worn, and human-carried (Buddemeir, Musolino, et al., 2016).
This study aims to develop an evaluation methodology for the comparison of radiation detection
and identification devices. This dissertation investigates the following research questions:
Question 1: What criteria should be used to determine the best device for a particular
operational scenario?
Question 2: What criteria should be used to determine the best portfolio of devices for a
given operational scenario?
Question 3: Is it best to seek a portfolio that consists of as many devices as possible, as
long as the cost does not exceed a given price threshold?
Question 4: When selecting a device portfolio that does not exceed a given price
threshold, how do we ensure that no device in the portfolio can be replaced
by another whose performance is better?
Intended Audience
This dissertation provides decisionmakers with an analysis tool for evaluating
radioisotope detection devices for use in detection, interdiction, and crisis response scenarios.
The implementation of the tool in specific applications can be described in terms of use cases, as
described in the sections below.

7
Use Case 1: Planning Activities
This tool can be used to help planners who are selecting equipment to establish, enhance
or sustain crisis response programs. The tool can be used to select a specific device or a portfolio
of devices in support of funding requests to government agencies that support the purchase of
radioisotope detection equipment. The planner would first determine the appropriate
organizational structure, staffing levels, risk factors, concept of operations and standard
operating procedures and that are appropriate for their jurisdiction. These factors inform the
number and types of devices that should be purchased. The planner would use the technique
described in Chapter 5 to determine the appropriate portfolio by inputting the number and type of
each device, along with budget constraints into the tool. This exercise may be repeated for
several funding profiles or for a variety of operational scenarios.
Use Case 2: PRND Training
Training coordinators provide technical advice and training to local, state, tribal and
federal emergency responders. They develop training exercises related to intentional as well as
accidental nuclear crises (U.S. Department of Energy Energy National Nuclear Security
Administration [NNSA] Administration Office of Nuclear Incident Response, 2019).
• Vignette 1: A training coordinator would use the tool functionality detailed in Chapter 4
to select the appropriate device for training PRND professionals in the use of a single
device as part a training course related to operations support or the use of a type of
detection device. As an example, a training coordinator developing a recertification
course in the use of human-portable backpack detectors would use the tool functionality
described in Chapter 4 to identify the device(s) to be used in the course. The trainer might
select two or three devices with the highest SMART scores. Similarly, a training program

8
designed to support commercial seaport operations might provide training on the
implementation of handheld radioisotope detection devices for secondary screening
operations. The trainer would select one or more top-performing RIIDs for use in this
training activity.
• Vignette 2: A training coordinator responsible for developing a large scale exercise
would select the appropriate portfolio size and composition based upon the command
structure of the agencies involved, concept of operations, operational scenario,
number/type of responders involved, etc. The planner would use the tool functionality
described in Chapter 5 to determine the most appropriate portfolio mix for a particular
operational scenario. As an example, an exercise simulating ground zero of nuclear
detonation site would require a portfolio consisting of personal radiation detectors
(PRDs), handheld radioisotope detection devices (RIIDs), and human portable backpack
radioisotope detection devices (BRDs). The training coordinator would use the tool
functionality described in Chapter 5 to select the appropriate devices for the portfolio.
Use Case 3: Community Policing
Local law enforcement personnel sometimes establish radiological safety programs in
conjunction with community partners (Mohres, 2020). Although these teams have no regulatory
authority, they may respond to incidents involving missing soil density gauges or radiography
cameras. Although the likelihood that these construction sources will be weaponized is low, it is
important ensure that they are properly secured and accounted for. A law enforcement
professional establishing a local radiological safety program would use the tool to identify the
appropriate portfolio of detection and protection devices.

9
Use Case 4: Federal Nuclear Incident Response
The Department of Energy’s Radiological Assistance Program (RAP) provides expertise
in nuclear weapon design, the characterization of nuclear and radiological materials as well as
the detection of radiological materials (U.S. Department of Energy Energy National Nuclear
Security Administration [NNSA] Administration Office of Nuclear Incident Response, 2019).
RAP provides federal, state, local, and tribal entities with first response capability in response to
radiological incidents; however, they do not preempt state, local, or tribal authorities. Their
standard response equipment includes devices with a combination of alpha, beta, gamma, and
neutron detection sensors. Typical missions include mobile surveys, chokepoint monitoring,
roving foot patrols, as well as aerial, venue and marine surveys. A RAP coordinator could use
the tool to identify portfolio(s) for each of these mission scenarios. The tool could be used to
identify the appropriate portfolio that corresponds to a particular mission or operational scenario.
As an example, a RAP coordinator could determine that a “We found a box” scenario that occurs
in a single building venue requires the support of a 5-person team equipped with five PRDs,
three BRDs, and five RIIDs.
Use Case 5: Military Response to International Crises
The Department of Defense (DoD) supports the Department of State (DoS) by serving as
lead agency in international Chemical Biological Radiological Nuclear (I-CBRN) crises in which
the host national (HN) requests the support of the U.S. Government (Department of Defense,
2016). The complexity of command relationships combined with the myriad of possible crises,
necessitates a structured approach to the selection of detection and identification equipment.
Personal Radiation Detectors (PRDs) may be used to support the commander’s priority of
ensuring that the exposure of all personnel is maintained as low as reasonably achievable

10
(ALARA). Handheld RIIDs and backpack devices will be needed in situations in which
nuclear/radiological sources must be located, detected or identified. Inter-agency teams could use
the tool to coordinate the appropriate ensemble of devices for response to international crises.
Use Case 6: EPA Radiological Response Team
The Environmental Protection Agency (EPA) designates on-site personnel to coordinate
preparedness planning and incident response to radiological releases at the local, national, and
international levels as outlined in the National Oil and Hazardous Substances Pollution
Contingency Plan (NCP).
The EPA’s Radiological Emergency Response Team (RERT) may provide radiation
monitoring, radionuclide analysis, as well as risk assessment in support of crises similar to the
2011 incident involving fires that threatened the Los Alamos National Laboratory. RERT
consists of a 20-member forward team and 15 support team members (Environmental Protection
Agency, 2017, p. 18). A RERT coordinator could use the tool to select the appropriate portfolio
of tools to support a 35-member RERT team.
Assumptions and Limitations of the Study
The purpose of this professional dissertation is to provide decisionmakers with a tool that
facilitates the selection of radioisotope detection devices for preventive radiological nuclear
detection (PRND) applications. The methodology demonstrates the application of a
multiattribute utility value model to the selection of radioisotope detection devices. The MAU
model parameters are primarily notional because they are based upon the input of a single
stakeholder. Ideally, the elicitation of judgments regarding scoring and weighting would be
performed by seasoned practitioners on a team of subject matter experts. An undertaking of this
magnitude is well beyond the scope of this project. Furthermore, linearity of value functions and

11
preference independence of the model criteria is assumed. The tradespace of candidate devices is
constrained by the availability of vendor data. This limitation could potentially impact the
outcome of the analysis.
Document Organization
The document is organized as follows:
• Chapter 2 presents a literature review of (a) the evolution of the radiological/nuclear
threat from the Cold War Era to the Post 9/11 threat landscape; (b) scientific foundations
of detection technology; (c) a discussion of the governance and policy issues that
contribute to the selection of detection devices; (d) the theoretical foundations of
multiattribute utility theory; and (e) the application of multiattribute utility theory in
portfolio optimization.
• Chapter 3 presents the risk and decision analysis methodology used in this study.
• Chapter 4 presents a methodology for evaluating radioisotope detection devices. It
investigates the Research Question 1: What criteria should be used to determine the best
device for a particular operational scenario?
• Chapter 5 presents a methodology for selecting the optimal portfolio of devices capable
of supporting detection, interdiction and crisis response missions. It investigates Research
Questions 2, 3 and 4:
o What criteria should be used to determine the best portfolio of devices for a given
operational scenario?
o Is it best to seek a portfolio that consists of as many devices as possible, as long as
the cost does not exceed a given price threshold?

12
o When selecting a device portfolio that does not exceed a given price threshold,
how do we ensure that no device in the portfolio can be replaced by another
whose performance is better?
• Chapter 6 summarizes the study and presents conclusions and possible extensions to this
project.

13
Chapter 2: Literature Review
Introduction
The foundation of this research is based upon four bodies of literature: (a) the American
response to the nuclear threat, (b) the scientific and technological underpinnings of nuclear
engineering, (c) a discussion of the governance and policy issues that contribute to the selection
of detection devices and (d) multiattribute utility theory (MAU).
Topic 1: History of the American Response to the Evolving Nuclear Threat
Nuclear threat reduction (NTR) refers collectively to the “integrated and layered
activities across the full range of U.S. government efforts to prevent and counter radiological and
nuclear threats” (Office of the Deputy Assistant Secretary of Defense for Nuclear Matters, 2020).
NTR activities support three primary goals: (a) preventing state both non-state actors from
acquiring nuclear weapons; (b) preventing the use of radiological and nuclear weapons; and
(c) minimizing the impact of radiological and nuclear event. It can be argued that NTR activities
preceded the development of the first atomic weapon. In 1939, Albert Einstein wrote President
Franklin Roosevelt to warn him of the dangers of atomic weapons (Einstein, 1939, as cited in
Roosevelt, 2020). President Roosevelt responded by establishing the initiative that is commonly
referred to as the Manhattan Project (Kelly, 2020; Rhodes, 1986; Roosevelt, 2020).
The Cold War, which began shortly after the end of World War II, was characterized by a
foreign policy of containment, also referred to as the Truman Doctrine (Beschloss, 2006), that
aimed to prevent the spread of communism (Gaddis, 2005). The U.S. offered economic, political,
and military support to nations at risk of authoritarian rule during the Cold War. In lieu of direct
conflict that could lead to their mutual destruction, the Superpowers engaged in a series of
regional “proxy wars” (Brands, 1993; Tonneson, 2010; Woodhouse, 2002).

14
The Cold War ended with the dissolution of the Soviet Union in 1991 (Clines, 1991).
Weapons-usable highly enriched uranium (HEU) and separated plutonium exist in hundreds of
locations, including the former Soviet Union, under varying levels of security. The logistics of
acquiring, building, and transporting a radiological dispersal device are significantly simpler than
those required to execute a nuclear terrorist attack involving a WMD. An IND can be constructed
from either highly enriched uranium (HEU) or plutonium. Guidelines developed by the
International Atomic Energy Agency (IAEA) indicate that as little as 55 pounds of HEU or 17
pounds of plutonium are required to manufacture a first-generation weapon. Because the
production of uranium enrichment requires resources that are only available to nation-states,
subversives must obtain HEU from existing stockpiles. Subversive groups may feel that the risk
and challenges associated with planning and executing an act of nuclear terrorism may be well
worth the proportional benefit derived from advancing their core objectives: revenge, renown,
and reaction (Richardson, 2006).
Several domestic and international programs have collaborated to protect the nation from
nuclear attack since World War II; they have collectively evolved into The Global Nuclear
Detection Architecture (GNDA). The GNDA consists of the intelligence, equipment, processes,
and agreements that protect the United States from such attacks. The objectives of the GNDA are
(a) to protect existing stocks of nuclear materials and weapons from diversion or theft; (b) to
enhance domestic interdiction and detection efforts; (c) to eliminate excess stocks of nuclear
materials; and (d) to detect the illicit movement of radiological or nuclear materials overseas.
The Domestic Nuclear Detection Office (DNDO) was established in 2005 within the Department
of Homeland Security to oversee the domestic portion of the GNDA. DNDO was tasked to
develop a layered strategy for nuclear detection (Government Accountability Office, 2019). In

15
2017, DHS initiated a reorganization that consolidated chemical biological radiological nuclear
and high yield explosives (CBRNE) counter-terrorism functions with the Office of Health
Affairs. Although the GNDA is no longer associated with a particular directorate, the GAO
reports that the responsibility for analyzing performance gaps and vulnerabilities will be
distributed throughout the Countering Weapons of Mass Destruction organization (Government
Accountability Office, 2019).
One objective of our study is to analyze the effectiveness of radioisotope detection
devices for use on health monitoring aspects of consequence management.
Topic 2: Science and Technology
Nuclear and Radiological Weapons
The prevention of insurgent groups from executing a terrorist attack using a radiological
or nuclear weapon is critical to our nation’s defense (Congressional Research Service, 2008).
Nuclear and radiological weapons are sometimes confused because they both contain radioactive
components. In truth, there are significant differences in component materials, processes and
blast effects (Knoll, 2010; Richardt et al., 2013).
Nuclear weapons were first used during World War II (Rhodes, 1986). Their design was
based upon fission, a nuclear reaction in which an atom’s nucleus is split into lighter nuclei
(Serber, 1992). Fission produces free neutrons and gamma photons. The term critical mass refers
to the amount of fissionable material necessary to generate a self-sustaining nuclear chain
reaction. The rapid assembly of a supercritical mass generates a nuclear explosion (Bethe &
Morrison, 2006). Implosion weapons generate explosions that are equivalent to hundreds of
kilotons of TNT. The resultant blast and heat effects can obliterate a large city and kill thousands
of people (Bernstein, 2008; Serber, 1992).

16
Thermonuclear fission weapons, which utilize a two-stage design, were developed during
the 1950s (Younger, 2009). The fission explosion that occurs during the first stage supplies the
energy required to trigger the weapon’s second thermonuclear stage, which causes the fusion of
deuterium or lithium, which in turn results in explosive yields between 1 and 100 megatons
(Bernstein, 2008).
Highly enriched uranium (HEU) and weapons grade plutonium (WGPu) are the principal
fissionable materials used to construct nuclear weapons. The procurement of the requisite special
nuclear material (SNM) required to construct an improvised nuclear device (IND) is a daunting
endeavor (Younger, 2009). Moreover, the construction of a nuclear weapon would require highly
specialized skills (American Institute in Aeronautics and Astronautics, 1993, 1998).
Conversely, the materials required to construct a radiological dispersive device (RDD), or
dirty bomb, are more easily acquired (von Winterfeldt & Rosoff, 2007). They comprise medical
or industrial radioactive materials as well as chemical explosives and aerosols to disperse the
radioactive materials. Although nuclear weapons can cause a massive number of human
casualties and significant damage to structures, the logistics associated with the design,
construction, procurement of materials, and transportation of a nuclear weapon are prohibitive to
terrorist organizations (Younger, 2009; Richardt et al., 2013).
RDDs are sometimes referred to as area denial weapons because they may contaminate a
several square mile area that must be evacuated and decontaminated. The diameter of the area
affected by the detonation of an RDD is dependent upon the direction that the wind is blowing
and the dispersal patterns of the radioactive particles (Philips et al., 2005). Although there would
be few, if any, casualties and minimal destruction in the event of an RDD attack, the derivative

17
economic and psychological effects would likely be significant (von Winterfeldt & Rosoff,
2007).
Detection of Illicit Radiological and Nuclear Materials
The ability to detect components of nuclear or radiological weapons is critical to defense
and homeland security operations. Radiation detection technologies serve dual functions: first to
intercept weapons or their components prior to an actual attack, and second, to support forensic
attribution functions after an attack (Philips et al., 2005).
The study of radiation detection requires an understanding of the fundamentals of nuclear
physics. Nuclear radiation, in the most general terms, involves the emission of energy as
particles or waves (Price, 1964). Radiation is best understood in terms of its interaction with
atoms, which are composed of a nucleus consisting of positively charged protons and neutrally
charged neutrons and an outer cloud of negatively charged electrons (Bethe & Morrison, 2006).
Radiation is classified as either ionizing or nonionizing, depending upon whether it has
adequate energy to attract electrons from atoms with which it comes into contact (Bethe &
Morrison, 2006). Atoms typically have an equal number of electrons and protons. However,
when an atom gains or loses an electron, it becomes charged and is referred to as an ion. In order
to restore its previous neutral charge, an ion seeks to bond with other charged particles. Non-
ionizing radiation is only as harmful as the amount of heat energy it transfers to an object; this is
the principle upon which microwave ovens operate. Conversely, overexposure to ultraviolet
light, another form of non-iodizing radiation, can increase an individual’s risk of cancer (Philips,
Nagel, & Coffey, 2005). Ionizing radiation represents a health threat because it can modify the
structure of an atom or even cause a mutation of a cell’s DNA molecules.

18
All nuclear detection technologies are designed to detect emissions from the decay of
radioactive nuclides, which can occur naturally, such as uranium and thorium, or are manmade,
such as plutonium and various fission products produced in a nuclear reactor (Knoll, 2010).
Alpha, beta, and neutron particles, as well as gamma and x-rays, are all forms of ionizing
radiation. These particles have attributes that affect their detectability; these attributes can also be
used to shield weapons from some types of detection instrumentation. The primary long-range
observables from nuclear materials are gamma rays and neutrons, which have mean free paths on
the order of 100 meters in air and only 10 cm in water (Congressional Research Service, 2008).
Neither artifact can be detected by satellite or high-altitude aircraft because their mean free paths
are only on the order of 100 meters (Congressional Research Service, 2008).
Background radiation contains both gamma rays and neutrons. Terrestrial, atmospheric,
and cosmic sources contribute to the natural gamma rays in the background environment,
referred to as naturally occurring radioactive material (NORM). The presence of neutrons in the
background is due to interactions between cosmic rays and the atmosphere as well as with large
objects such as buildings. A nuclear or radiological weapon may not always be detectable above
the natural background; detectability is influenced by the configuration of the material, the type
of detector, the amount of shielding, and distance from the source (Philips, Nagel, & Coffey,
2005).
Topic 3: Policy Constraints that Impact the Selection of Instrumentation
An analysis of the applicability of radioisotope detection equipment for use in
consequence management (CM) missions is critical to the implementation of the new policy. An
analysis of single radiological events can be performed in order to determine the resources
required to support a radiological event. The primary event types are:

19
• Domestic nuclear explosion
• Nuclear power plant incident/event involving a significant release
• Use of alpha RDD/failed IND
• Beta-Gamma radiological dispersal device
Devices are evaluated against the following mission areas, as detailed in Lawrence
Livermore National Laboratory (2017):
• Cold zone missions
o Worker exposure control
o Worker dose monitoring
o Radiation survey
o Person/object external contamination detection
o Isotope identification
• Hot zone missions
o Worker exposure control
o Worker dose monitoring
o Radiation survey (limited to hot zone)
• Dangerous radiation zone missions
o Worker exposure control and monitoring
Although detection devices were specifically designed for R/N counter-terrorism
missions, they may possess the functionality to support all or a subset of CM operations. The
trade space is wide: There are hundreds of radioisotope detection and/or identification
instruments, each with its own attributes and performance levels. Tables 1and 2 summarize the

20
applicability of the instrument archetypes for use in detection, interdiction and crisis response
and missions (Lawrence Livermore National Laboratory, 2017).

21
Table 1.
Body-Worn Devices
Category Defining Characteristics Mission Applicability
Personal radiation
detector (PRD)
Highly sensitive, can detect small
changes from background.
Alarming, body worn device
capable of passing low exposure
rate tests of ANSI N42.32.
Typically uses scintillation
detectors.
PRND: Detection of low-level
radiation for contraband
investigation.
CM: Environmental and
personnel contamination
surveys in cold zone.
Spectroscopic
personal radiation
detector (SPRD)
Highly sensitive, can detect and
identify low levels of radiation.
Alarming, body worn device
capable of passing low exposure
rate tests of ANSI N42.48.
Typically uses scintillation
detectors.
PRND: Detection and
identification of low-level
radiation for contraband
investigation.
CM: Environmental and
personnel contamination
surveys in cold zone,
radionuclide ID.
Extended range
personal radiation
detector (ER-PRD)
Highly sensitive, can detect small
changes from background.
Alarming, body worn device
capable of passing low exposure
rate tests of ANSI N42.32.
Extended range, with the capability
to measure up to 10 R/h or more.
PRND: Detection of low-level
radiation for contraband
investigation.
CM: Cold and hot zone
survey and responder
exposure control.
Personal emergency
radiation detectors
(PERD) & monitors
High range, alarming, body worn
device capable of operating above
10 R/h, potentially up to 1,000 R/h
(ANSI N42.49A).
Capable of operating in harsh
environments.
CM: Detection and entry into
hot zone, exposure control
and possibly dose monitoring
tool.
Electronic personal
dosimeter (EPD)
High range, alarming, body-worn
device for occupational workers to
measure personal dose equivalence
for regulatory compliance.
Performance requirements can be
found in ANSI N42.20.
CM: Hot zone detection,
responder exposure control
and dose monitoring tool if
ruggedized.
Note: Adapted from Mission Analysis for Using Preventive Radiological/Nuclear Detection
Equipment for Consequence Management by Lawrence Livermore National Laboratory,
2017, pp. 3-4. https://doi.org/10.2172/1515353.

22
Table 2.
Human-Carried Devices
Category Defining Characteristics Mission Applicability
Radioisotope
identification device
(RIID)
Hand-held devices that detect low
levels of radiation up to 10 milliR
per hour (mR/h) or more.
PRND mission related performance
requirements in ANSI N42.33.
Other requirements in N42.17A
(normal conditions) and N42.17C
(extreme conditions).
PRND: Detection of low-
level radiation for
contraband investigation.
CM: Workplace or public
safety. Can be used to find
contamination or hotline.
Hand-held survey
meter: Low range
Hand-held devices that measure
high radiation levels to 10 R/h.
Requirements can be found in
N42.17A (normal conditions) and
N42.17C (extreme conditions).
PRND: Detection of low-
level radiation for
contraband investigation.
CM: Workplace or public
safety. Can be used to find
contamination or hotline.
Hand-held survey
meter: High range
Hand-held devices that measure
high radiation levels to 10 R/h.
Requirements can be found in
N42.17A (normal conditions) and
N42.17C (extreme conditions).
CM: Detection and entry
into hot zone and responder
exposure control.
Human-portable
detector (backpack)
Very sensitive radiation detectors.
Large (backpack or suitcase sized).
Capable of passing radiological
performance tests indicated in
ANSI
N42.43.
PRND: Detection of low-
level radiation for
contraband investigation.
CM: Environmental and
personal contamination
surveys in cold zone.
Note: Adapted from Mission Analysis for Using Preventive Radiological/Nuclear Detection
Equipment for Consequence Management by Lawrence Livermore National Laboratory,
2017, pp. 3-4. https://doi.org/10.2172/1515353ary of
Topic 4: Multiattribute Utility (MAU) Methodology
The foundations of decision analysis were established in the 18th century by Abraham de
Moivre (1718) and Thomas Bayes (1763). DeMoivre proposed a frequency approach to statistics
while the Bayesian approach represented a departure from the conventional way of thinking
about event probabilities. Rather than reasoning about what could happen from a given state, he
considered the events that would lead to a given state.

23
Decision analysis (DA) is a prescriptive theory comprised of models and tools that can be
applied to a broad range of complex decision problems (Edwards et al., 2007). Daniel Bernoulli
(1954) laid the foundation for expected utility theory when he proposed that decisionmakers
should (a) consider the likelihood of the possible outcomes associated with a decision and (b)
quantify the desirability of those outcomes. Von Neumann and Morgenstern (1944) introduced
the concept of rational decisionmaking, which posits that a rational person should use a utility
measure rather than willingness-to-pay in scenarios characterized by uncertainty.
Robert Schlaifer (1951) presented a practical approach to the application of Bayesian
decision theory to management problems. Howard Raiffa and Robert Schlaifer (1961) co-
authored Applied Statistical Decision Theory, in which they proposed an analytical decision-
making technique that expresses the decisionmaker’s preferences in terms of numeric utilities.
The possible outcomes of the decision are assigned weights, which are expressed as statistical
probabilities.
Raiffa (1968) published one of the first books on decision analysis. In 1976, Raiffa and
Keeney published the seminal work on multiattribute utility theory (MAUT; Keeney & Raiffa,
1976). MAUT recognizes that complex decision problems are characterized by multiple
objectives, some of which are conflicting. As a result, the achievement of one objective may
occur at the expense of another (von Winterfeldt & Edwards, 1986). This requires the
decisionmaker to consider and quantify tradeoffs associated with varying degrees of achievement
associated with conflicting objectives. The uncertainty associated with most complex decision
problems further complicates the process. The approach proposed by Keeney and Raiffa (1976)
is the basis of the methodology implemented in this work. The steps include:

24
• Structuring the problem: Articulation of the decisionmaker’s objectives and measures of
performance.
• Identification and quantification of uncertainty: This step involves identifying the
consequences associated with each alternative, including a probability distribution
representing the likelihood of consequences associated with each alternative.
• Quantification of Preferences: This step involves the assessment of the decisionmaker’s
utility function. The utility function is a vector that represents the levels associated with
attributes.
• Evaluation of alternatives: The expected utility of each alternative is calculated during
this step. Sensitivity analysis is performed to gain insight into how varying parameters
associated with the probability distributions and utility function impact the expected
utility of each alternative.
Detlof von Winterfelt and Ward Edwards (1986) elucidated the link between decision
analysis and behavioral analysis, a subdiscipline of cognitive psychology. An important
contribution of this work is a discussion of the distinction between utility and value. In decision
theory, value represents a transformation on a physical scale while utility is actually a
transformation performed on value; utility captures the decisionmaker’s attitude toward risk.
The methodology used in this study applies the simple multiattribute rating technique
(SMART), a version proposed by Edwards (1971, 1977) based on an additive model. The
SMART swing weighting procedure (Keeney & von Winterfeldt, 2011) is used to elicit
preferences from a PRND stakeholder.
MAUT has been applied to decision problems, including sustainable energy decisions
(Nikou & Klotz, 2014), transportation systems (Yin, 1981), medical application software (Swan,

25
2004), infrastructure design (Zavadskas & Vaidogas, 2009), the selection of scholarship
recipients (Yeh, 2003), nuclear power plant safety (Beaudouin, 2015), evaluation of plutonium
disposition options (Edwards et al., 2007), and portfolio optimization (Ehrgott et al., 2009).
Keeney and von Winterfelt provide a step-by-step application of MAUT to the development of a
value model for homeland security applications (Keeney & von Winterfeldt, 2011).
Portfolio decision analysis applies tools such as influence diagrams, decision trees, value
models, and utility functions to the problem of resource allocation (Salo et al., 2011). The task of
eliciting MAU model parameters relative to portfolio decision problems is complex and can be
fraught with errors (Salo & Liesio, 2006; Sharpe & Keelin, 1998). Fasalo et al. (2011) discussed
the cognitive processes that influence how decisionmakers approach resource allocation.
Portfolio decision problems sometimes require decisionmakers to select a subset of
alternatives based upon a goodness measure that may vary based upon scenario. One approach is
to implement an evaluation function that captures the decisionmaker’s preferences (Farquhar &
Rao, 1976) relative to achieving a balanced portfolio. In some cases, balance refers to the need to
achieve homogeneity of some attributes and heterogeneity of others.
Some portfolio decision problems involve the selection of individual as well as groups of
alternatives. Cardinal et al. (2011) presented a student selection case study that included both
individual and group criteria. Individual criteria referred to the quality of the candidate’s
application while group criteria included factors such as gender balance, distribution among
professional tracks, and so forth. Similarly, the MAU decision problem under consideration
encompasses both the selection of an individual device as well as an ensemble of devices.
Chapter 4 presents the criteria for selecting individual devices; these are benefit and cost data
provided by the vendors. Chapter 5 presents a value model for selecting a portfolio of devices.

26
Chapter 3: Methodology
The detection of the contraband components used to construct a radiological or nuclear
device can be performed using a variety of equipment types: devices capable of actively
interrogating cargo containers, passive detection equipment used for scanning smaller containers,
handheld units used to inspect parcels, as well as the neutron and gamma-ray detectors utilized at
ports of entry.
First responders often rely on human-portable devices. There are three primary types of
portable devices. Personal radiation detectors (PRDs) often include dosimeters, handheld
radioisotope identification devices (RIIDS), and backpack radioisotope detectors (BRDs).
The problem of evaluating radiological/nuclear detection technologies will be structured
as a multiattribute utility model (MAU) model. This chapter will demonstrate the development of
a simple MAU model for the evaluation of a notional set of backpack radiation devices (BRDs).
Development of MAU Model
The development of a multiattribute utility (MAU) model requires the specification of
alternatives, objectives, and performance measures (Eisenfuhr et al., 2010). The MAU model
will be developed as described in von Winterfeldt and Edwards (1986, p. 273), which involves
the following steps:
1. Definition of objectives, attributes and alternatives
2. Evaluation of alternatives relative to attributes
3. Determination of attribute weights
4. Evaluate alternatives using attributes and weights
5. Perform sensitivity analysis
6. Communicate findings and make recommendations

27
Definition of Objectives, Attributes, and Alternatives Research Methods
The objectives and attributes were obtained from informal discussions with first
responders. The first responders were members of regional law enforcement and firefighting
teams. The discussions were conducted in conjunction with a review of vendor literature.
Objective 1: Maximize Detection Accuracy
A key measure of system performance is the detector’s ability to detect a nuclear or
radiological weapon properly. The type of materials detectable by the device is also a key
performance factor. The types of threat objects in the potential trade space is large and can be
initially described by the type of radiological or nuclear material, the mass of the material
present, and the presence of masking or shielding. During this phase of the study, this
constructed measure will reflect whether the device is capable of detecting and identifying
special nuclear material (SNM), naturally occurring radioactive material (NORM), and industrial
sources and medical sources. The attributes corresponding to Objective 1 are:
• A
1
: Detection rate (%) – the detector’s ability to properly detect the contraband material
used to build a nuclear or radiological device.
• A
2
: False alarm rate (numeric) – This definition has been adapted from the Domestic
Nuclear Detection Office (2013, p. 17). The false alarm rate is defined as the number of
alarms generated within a 10-hour period in an area with a controlled background.
According to the Domestic Nuclear Detection Office (2013), an acceptable false alarm
rate does not exceed five alarms/identifications within the test period.
• A
3
: Threat library – This constructed measure indicates whether a device is capable of
detecting and identifying special nuclear material (SNM), naturally occurring radioactive
material (NORM), or industrial sources and medical sources.

28
o User definable superset of ANSI 42.53 (100)
o ANSI 42.53 (75)
o Subset of ANSI 42.53 (40)
Objective 2: Maximize Operational Effectiveness
Operational effectiveness encompasses the hierarchy of performance factors depicted in
Figure 2. Because communication reachback, data collection, form factor, alarm type resilience,
and environmental conditions are addressed by the IEEE Standards Association (2013), our
analysis will focus on battery life, regulatory compliance, and weight. The attributes
corresponding to Objective 2 are:
• A
4
: Battery life – the amount of time that the device can operate continuously before
having to replace the batteries.
• A
5
: Regulatory compliance (yes/no) – refers to the device’s compliance with
ANSI/IEEE N42.53 ‘American National Standard Performance Criteria for
Backpack Based Radiation Detection Systems Used for Homeland Security’.
• A
6
: Weight (pounds) – this attribute refers to the weight of the device, inclusive of
the battery and the backpack or other form-factor (vest, briefcase).

29
Figure 2.
Hierarchy of Operational Attributes

Objective 3: Minimize Cost
This objective refers to the purchase price and annual maintenance, as well as any
training costs associated with the purchase of the device. This objective is associated with a
single attribute, cost (Attribute A
7
), which is measured in thousands of U.S. dollars.
Evaluation of Alternatives Relative to Attributes
The MAU methodology will be demonstrated on a set of notional backpack radiation
detection (BRD) devices, referred to as A, B, and C, that have been assigned the values in
Table 3. This notional example demonstrates the MAU methodology. Device A is a lightweight
backpack that is used to provide just-in-time training for first responders; it must be calibrated by
the manufacturer. Device B is a rugged unit that has been used to provide just-in-time training
for new responders. Device C provides real-time nuclide identification but does use proprietary
Operational
Effectiveness
flexible form
factor
vest overlay
backpack
briefcase
communication
reachback
types of alarm
audio
tactile
visual
data collection resilience
water
resistance
dust resistance
drop resistance
shock
resistance
dimensions
environmental
conditions
temperature
humidity
rain

30
parts and must be calibrated by the manufacturer. Moreover, the global positioning system
location and alarm data have not been paired to support mapping capability.
Table 3.
Summary of Attribute Measures
Objective Measures (units) Alternatives
Device A Device B Device C

Benefit
attributes

Detection
accuracy
Detection
rate as proxy
Numeric (%) 94 90 98
False alarm
rate as proxy
Numeric
# false alarms in
10-hour period
4 6 13
Threat
library
Constructed
User definable
superset of ANSI
42.53 (100)
ANSI 42.53 (75)
Subset of ANSI
42.53 (40)
100 40 75
regulatory compliance Yes/no
Compliant with
ANSI 42.53 (100:
yes)
Not compliant
with ANSI 42.53
(0: no)

Yes

No Yes

Weight Pounds 9 13 20
Battery life Hours 24 65 20
Cost Purchase price and annual
maintenance
Thousands of
dollars
22 34 53

31
Determination of Attribute Weights
The value assigned to each device is determined by the formula:

(1)
where:
• v(a
j
) refers to the value assigned to alternative j
• w
i
refers to the weight assigned to attribute i (there are a total of n attributes)
• x
ij
refers to the value of attribute i assigned to alternative j
Table 4 identifies x
ij
for the n = 7 attributes in our notional example. The weights (w
j
) are
determined using the SMART swing weight procedure (Clemen & Reilly, 2001). The SMART
methodology makes no distinction between values and utilities. Table 5 reflects the use of values
in the swing weight procedure. These steps will be performed on the non-cost related attributes.
The utility of the cost attribute will be determined separately.
Table 4.
Best–Worst Attribute Values
Best Worst
Detection rate 90 98
False alarm rate
13 4
Threat objects 40 100
Regulatory compliance No Yes
Battery life 20 65
Weight 20 9

32
Swing Weight Procedure
1. Create hypothetical alternatives with only worst and only best levels
• Worst device (detection rate = 90, false alarm rate = 13, threat objects = 40,
regulatory compliance = no, battery life = 20, weight = 20)
• Best device (detection rate = 98, false alarm rate = 4, threat objects = 100, regulatory
compliance = yes, battery life = 65, weight = 9)
2. Determine which attribute you would first like to change (“swing”) from worst to best
• #1: threat object
3. Determine which attribute is second, third, etc. (Rank Weight column in Table 3)
• #2: detection rate
• #3: false alarm rate
• #4: regulatory compliance
• #5: battery life
• #6: weight
4. Assign 100 points to first “swing” attribute (Rated Weight column in Table 5)
5. Assign relative numbers to other preserving swing ratios (Rated Weight column in
Table 5)
6. Normalize to add to 1 (Normalized Weight column in Table 5)

33
Table 5.
Determination of Attribute Ranks and Weights
Range
Rank
Weight
Rated
Weight
Normalized
Weight Worst Best
Detection rate 90 98 2 80 0.21
False alarm rate (#
false alarms in 10-
hour period)
13 4 3 80 0.21
Threat objects 40 100 1 100 0.26
Regulatory
compliance
0 100 4 50 0.13
Battery life (hours) 20 65 5 40 0.11
Weight (pounds) 20 9 6 30 0.08
Sum 1.0

Evaluate Alternatives using Attributes and Weights
The swing weight procedure described in Chapter 3 is used to assign ratings and weights
to objectives. Benefit, or non-cost related, objectives are analyzed separately from the cost-
related objective.
Evaluate Alternatives using Attributes and Weights for non-Cost Related Objectives
The utilities corresponding to the alternatives are presented in Table 6 The assigned
utilities were chosen to ensure the linearity of the model. A value of 100 is assigned to the
attribute corresponding to alternative corresponding to the best-performing attribute and 0 is
assigned to the worst-performing alternate. The remaining utility is proportional to the ratio of its
magnitude to the length of the interval defining the range of the attribute’s values. The SMART

34
model makes no distinction between value and utility. Ideally, the weights and values would be
generated based upon the input of a subject matter expert (SME). In the absence of such input,
values corresponding to similar devices will be used.
Table 6.
Multiattribute Utility Model
Criteria
Device
A
Device
B
Device
C Rank
Rated
Weight
Normalized
Weight (w
i
)
Detection rate 0 50 100 2 80 0.21
False alarm rate 100 0 78 3 80 0.21
Threat objects 100 0 35 1 100 0.26
Regulatory compliance 100 0 100 4 50 0.13
Battery life (hours) 9 100 0 5 40 0.11
Weight (pounds) 100 64 0 6 30 0.08
Sum 409 214 313 380 1.00
unweighted MAU
= sum / 6 (# attributes) 68 36 52

weighted MAU
𝒖𝒖 � 𝒂𝒂 𝒋𝒋 � = ∑ 𝒘𝒘 𝒊𝒊 𝑢𝑢 𝟔𝟔 𝒊𝒊 = 𝟏𝟏 � 𝒙𝒙 𝒊𝒊 𝒋𝒋 �
69 16 60

Evaluate Alternatives using Attributes and Weights for Cost Attribute
This section presents a cost-inclusive analysis of the three alternative devices. This
process begins with a price vs. utility study. The value corresponding to utility is obtained from
the value corresponding to price on the last row of the weighted MAV in Table 6. Price is
presented using negative values to facilitate the analysis. Table 7 presents the utility values and
the price of each device.

35
Table 7.
Utility vs. Price

Utility
(from last row in Table 6) Price (K)
Device A 69 22
Device B 16 34
Device C 60 53

Figure 3.
Utility vs. Cost

Figure 3 presents a plot of the price vs. utility data from Table 7. Device A dominates the
other two devices because it corresponds to a point that is higher, indicating a lower cost, and
farther right, indicating a larger utility than Devices A and B.
22, 69
34, 16
53, 60
0
10
20
30
40
50
60
70
80
0 10 20 30 40 50 60
Utility
Cost
Utility vs. Cost
A
B
C

36
Figure 4.
Sensitivity Analysis (Utility vs. Cost Weight)

The sensitivity analysis presented in Figure 4 indicates that Device A dominates the other
two devices throughout the range of weights applied to the device cost. According to this
analysis, the decision maker should always select Device A.
We now complete the MAU model by determining the rank and rated weight
corresponding to the cost attribute. The rated weight corresponding to device cost is determined
as follows:
1. We determine the rated weight of the cost attribute by creating fictional alternatives that
correspond to the worst and best options for weight and cost.
• Worst: (20 lbs., 53K)
• Best: (9 lbs., 22K)
The difference between the two options corresponds to an 11-pound difference in device
weight and a 31K difference in device cost. Suppose that the 31K cost reduction is preferable to
0.00
10.00
20.00
30.00
40.00
50.00
60.00
70.00
80.00
90.00
100.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00
Utility
Weight of Device Cost
Sensitivity Analysis (Utility vs. Cost)
Device A Device B Device C

37
the 11-pound reduction in device weight. The weight rating for the cost attribute is determined
by quantifying the amount of preference. If the SME determined that it is twice as valuable, then
the cost attribute is assigned a rated weight that is double that of the device weight attribute (2 x
30 = 60). The MAU is updated to reflect that the rated weight of the cost attribute is 60.
2. We now determine the weight of the cost attribute in order to determine the weighted
MAU scores.
Sum the rated weights for all attributes:
 80 + 80 + 100 + 50 + 40 + 30 + 60 = 440
3. Determine the new weight of each attribute by dividing the attribute’s utility by the sum
obtained in Step 2.
4. Determine the weight of the cost attribute by subtracting the sum obtained in Step 3 from
1 (1 − 0.86 = 0.14).
Table 8 displays the revised MAU.

38
Table 8.
Completed MAU
Criteria Device A Device B Device C Rank
Rated
Weight
Normalized
Weight
Detection rate 0 50 100 2 80 0.18
False alarm rate 100 0 78 3 80 0.18
Threat library 100 0 35 1 100 0.23
Regulatory
compliance
100 0 100 5 50 0.11
Battery life
(hours)
9 100 0 6 40 0.09
Weight
(pounds)
100 64 0 7 30 0.07
Subtotal (non-
cost attributes)
409 214 313 0.86
Price 100 80 0 4 60 0.14
Total 509 294 313 440 1.0
Unweighted 73 42 45
Weighted 73.81 33.68 51.09

Summary
This chapter demonstrated the use of a MAU model for the evaluation of a small set of
backpack mounted radioisotope identification devices. The attributes were relative to the R/N
detection mission only. The analysis performed in the next chapter will also encompass the crisis
response missions.

39
Chapter 4: Evaluation of Backpack Radiation Detectors
This chapter demonstrates the application of the multiattribute utility methodology in the
evaluation of radioisotope detection devices. This chapter is responsive to the research question
“What criteria should be used to determine the best device for a particular operational scenario?”
The analysis performed in this chapter differs from that performed in Chapter 3 in that the model
alternatives correspond to actual devices. The analysis focuses on five backpack radiation units:
Mirion Technologies SPIRPack–G, Berkeley Nucleonics SAMPACK, Bubble Technology
Industries FlexSpec, Thermo-Scientific’s PackEye GN-2, and Bruker Sentry.
The model parameterization is based upon the input of a single stakeholder rather than on
the opinions of a team of subject matter experts. The resulting analysis is based upon notional
values for the attribute weights and rankings. Notional values were used for detection rate and
false alarm rate attributes because these values are not publicly available.
The stakeholder has worked with a preventive radiological nuclear detection (PRND)
team for 25 years. He has provided expertise to other PRND teams on the selection of
radioisotope detection devices. Our communications consisted of telephone calls, email
messages, and responses to questions regarding the model’s objectives, attributes, scores, and
weights.
The analysis follows the same methodology described in Chapter 3. In response to input
provided by a PRND stakeholder, the set of model objectives has been expanded to support crisis
response activities. The new objective, to maximize support of post-event activities, refers to
functions that must be performed after a radiological or nuclear event.
The primary metric for measuring the efficacy of radioisotope detection/identification
equipment in post-event activities is the device’s operational range. Operational range refers to

40
the level of radiation that a device can reasonably measure. Devices used in interdiction and
detection missions are most effective if they are sensitive enough to detect/identify contraband at
low levels or when masked or shielded by other substances. Conversely, devices used in crisis
response missions should be able to identify high levels of radiation.
Areas impacted by radiological/nuclear events are categorized by increasing radiation
level: cold zone, hot zone, and dangerous zone. The cold zone is characterized by elevated
radiation levels that do not require access restrictions. In the cold zone, backpack radiation
detectors can be used to support consequence management missions involving workplace or
public safety (Lawrence Livermore National Laboratory, 2017, p. 18). The post-event support
attribute assumes four possible values:
• Full (100) – The device provides adequate support for the mission
• Marginal (75) – The device meets the minimum requirement for the mission
• Insufficient (25) – The device does not meet the minimum requirements for the mission
• Information not specified (0)
Following the convention identified in Lawrence Livermore National Laboratory (2017,
p. 16), a device’s rating will correspond to the lowest rating for the mission. For example, a
device is scored as below in the cold zone will be assigned an overall mission value of
insufficient.
The process for developing an MAU decision model begins with the development of a
consequence table that contains the objectives, measures relative to the alternatives under
consideration. The objectives presented in Table 9 represent the criteria that will be used to select
the best device for detection, interdiction, and crisis response scenarios (Question 1).

41
Table 9.
Consequence Table
Objectives Measures Alternatives
SpirPack SAMPack FlexSpec PackEye Sentry
Detection
rate
% 93 90 92 94 96
False alarm #/10-hour
period
7 8 9 10 5
Isotope
Identification
Categorical ANSI +
Vendor
ANSI +
Vendor
ANSI +
Vendor
Not
specified
ANSI +
Vendor
Threat
notification
mode
Categorical Two All 3 All 3 Not
specified2
Two
Response
time
Seconds 2 2 Not
specified
1 2
Weight Pounds 9.9 9 15 16.5 14
Battery life Hours 40 8 12 30 8
Post-event
support
Constructed 25 75 75 50 50
Price USD
(thousands)
16.2 25 30 25 25

Evaluation of Alternatives Relative to Attributes
The attribute ratings in the consequence table are converted to single attribute values on
the same scale to facilitate comparisons. Each attribute rating is assigned a value between 0 and
100. The best attribute in a category is assigned a value of 100, and the worst value is assigned a
value of 0. The remaining ratings are assigned values within the range. Our ratings were
informed by the subjective input of a PRND stakeholder. Because the ratings are based upon the

42
input of a single stakeholder rather than that of a team of subject matter experts, the ratings are
considered notional. These mappings are detailed in Table 10
Table 10.
Conversion of Consequences to Single Attribute Values
Objectives Measures Alternatives
SPIRPack SAMPack FlexSpec PackEye Sentry
Detection
rate
% 30 0 20 50 100
False alarm #/10-hour
period
40 35 30 25 100
Isotope ID categorical 100 100 100 0 100
Threat
notification
mode
categorical 75 100 100 100 50
Response
time
seconds 75 75 0 100 75
Weight pounds 90 100 20 0 25
Battery life hours 100 0 40 80 0
Post-event
support
categorical 0 100 100 50 50

Calculation of Weights using Swing Weight Procedure
Consequence tables present the facts regarding each candidate device. The weights
assigned to each attribute represent value tradeoffs. The generation of a consequence table is
ideally performed with input from one or more subject matter experts (SMEs) while the
determination of attribute weights is based upon stakeholder judgments. A stakeholder with more

43
than twenty years of experience in PRND operations provided input to the swing weight
procedure.
The swing weight procedure was used to assign weights to each parameter as described in
Chapter 3. The swing weight procedure is based upon a set of N + 1 fictional alternatives; N
corresponds to the number of attributes. One alternative is parameterized using the “worst-case”
value of each attribute. The remaining N alternatives are parameterized using the worst-case
value of all except one attribute; that attribute is parameterized using the best outcome. The
PRND stakeholder rank-ordered the N + 1 alternatives by successively selecting the alternative
corresponding to the attribute deemed most important. At the end of the process, the worst-case
alternative was ranked N + 1 and the most valued alternative w ranked first.
Table 11 details the results of the swing weighting procedure for non-cost related
attributes. The non-cost related objectives are analyzed separately from the cost-related
objectives. Rates and ranks were assigned to each objective. The swing weight procedure
generated the values corresponding to the Rank Weight column. The values corresponding to the
column labeled rated weight were generated by having the stakeholder assign a value between 0
and 100 to each objective, based upon the ranking of the importance of the value difference
between the worst and the best consequence. The sum of the rated weights is used to determine
the normalized weight. The normalized weight of an objective is obtained by dividing the
objectives rated weight by the sum of all rated weights. For example, the normalized weight of
detection rate is equal to 100 / 475 (21.05).

44
Table 11.
Calculating Weights for Non-Cost Related Attributes
Objectives Measures Worst Best
Rank
Weight
Rated
Weight
Normalized
Weight
Detection
rate
% 90 96 1 100 21.05
False alarm #/10-hour
period
10 5 2 90 18.95
Isotope
Identification
Categorical 1 4 3 70 14.74
Threat
notification
mode
Categorical 0 3 5 50 10.53
Response
time
Seconds 10 1 4 60 12.63
Weight Pounds 16.5 9 6 45 9.47
Battery Hours 8 40 8 20 4.21
Post-event
support
Categorical 25 75 7 40 8.42
TOTAL 475 100.00

Normalized weights are used to calculate the weighted multi-attribute values (MAV) for
each alternative. The utility value for each non-cost related objective is calculated by summing
the product of the normalized weight and the value assigned to each objective as described in
Chapter 3:
(2)

45
In this equation, w
i
is the normalized weight associated with objective i and v
i
is the value
assigned to objective i for alternative j. Using this formula, the weighted utility for SPIRPack is
calculated as indicated in Table 12.
Table 12.
Calculation of MAV for SPIRPack Device
SPIRPack
Objective Value
Normalized
weights
[(Single attribute value) *
(Normalized weight)] / 100
Detection rate 30 21.05 6.32
False alarm 40 18.95 7.58
Isotope Identification 100 14.74 14.74
Threat notification mode 75 10.53 7.9
Response time 75 12.63 9.47
Weight 90 9.47 8.52
Battery 100 4.21 4.21
Post-event support 0 8.42 0
Total 510 100 58.74
Unweighted MAV = 510 / 8 = 63.75 Weighted MAV = 58.74

Table 13 contains the unweighted and weighted MAV scores for all devices.

46
Table 13.
Calculating Utility Values for Non-Cost Related Objectives
Objectives
SPIRPack
𝑣𝑣 𝑖𝑖
SAMPack
𝑣𝑣 𝑖𝑖
FlexSpec
𝑣𝑣 𝑖𝑖
Packye
𝑣𝑣 𝑖𝑖
Sentry
𝑣𝑣 𝑖𝑖
Normalized
weights
𝑤𝑤 𝑖𝑖
Detection rate 30 0 20 50 100 21.05
False alarm 40 35 30 25 100 18.95
Isotope
Identification
100 100 100 0 100 14.74
Threat
notification
mode
75 100 100 100 50 10.53
Response time 75 75 0 100 75 12.63
Weight 90 100 20 0 25 9.47
Battery 100 0 40 80 0 4.21
Post-event
support
0 100 100 50 50 8.42
Total 510 510 410 405 500 100.00
Unweighted
utility = (total /
#attributes)
63.75 66.88 51.25 50.63 62.50
Weighted utility
values
58.74 59.26 47.16 41.23 70.75

Table 14 presents the rankings of the alternatives. SAMPack is ranked first using the
unweighted approach; it is ranked second using the weighted MAV. Although Sentry is ranked
first using the weighted approach, it is ranked third using the unweighted approach. SPIRPack is
ranked second using the unweighted approach and third using the unweighted approach.
FlexSpec and PackEye are ranked fourth and fifth, respectively, using both approaches. A

47
decisionmaker could be confident in selected SAMPack because it is not ranked lower than
second place using either approach. Similarly, the decisionmaker could confidently eliminate
FlexSpec and PackEye from a tiered ranking of alternatives since they consistently ranked fourth
and fifth place using both approaches.
Table 14.
Unweighted vs. Weighted MAV (for Benefit Objectives)
Ranking Unweighted MAV Weighted MAV
1 SAMPack (66.88) Sentry (70.75)
2 SPIRPack (63.75) SAMPack (59.26)
3 Sentry (62.50) SPIRPack (58.74)
4 FlexSpec (51.25) FlexSpec (47.16)
5 PackEye (50.630 PackEye (41.23)

The next phase of the analysis involves the determination of the overall value for benefit
and cost attributes. The PRND stakeholder provided input regarding the weights assigned to the
non-cost attributes. A slightly different procedure is used to determine the weight to be assigned
to the cost-related attribute. The stakeholder is asked how much they would be willing to pay to
swing the price of the most expensive device from worst to best. In this instance, the stakeholder
was willing to pay double in exchange for a reduction in price from $30,000 to $16,200. This
means that the corresponding rating assigned to that attribute would be ½ = 0.50. Table 15
reflects the values assigned to the price objective for each alternative.

48
Table 15.
Assignment of Values for BRD Price Using Swing Weighting Method
Objectives Measures Alternatives
SPIRPack SAMPack FlexSpec PackEye Sentry
Detection rate % 30 0 20 50 100
False alarm #/10-hour
period
40 35 30 25 100
Isotope
Identification
Categorical 100 100 100 0 100
Threat
notification
mode
Categorical 75 100 100 100 50
Response time Seconds 75 75 0 100 75
Weight Pounds 90 100 20 0 25
Battery life Hours 100 0 40 80 0
Post-event
support
Categorical 0 100 100 50 50
Price USD
(thousands)
100 25 0 25 25

Table 16 presents the results of the calculation of weights for the cost-related objective.
The rating of 50 is assigned to the rated weight of the price objective; this corresponds to a rated
weight of 6. The normalized weight for each objective is recalculated using the new rated weight
total of 525.

49
Table 16.
Weights Assigned to Benefit and Cost Objectives
Objectives Measures Worst Best
Rank
Weight
Rated
Weight
Normalized
Weight
Normalized
Weight
(with price)
Detection
rate
% 90 96 1 100 21.05 19.05
False alarm #/10-hour
period
10 5 2 90 18.95 17.14
Isotope
Identification
Categorical 1 4 3 70 14.74 13.33
Threat
notification
mode
Categorical 0 3 5 50 10.53 9.52
Response
time
Seconds 10 1 4 60 12.63 11.43
Weight Pounds 16.5 9 7 45 9.47 8.57
Battery Hours 8 40 9 20 4.21 3.81
Post-event
support
Categorical 25 75 8 40 8.42 7.62
30 16.2
Total 475 100.00 90.48
Price USD
(thousands)
30 16.2 6 50 0 9.52
Total
(including
price)
525 100

The new normalized weight that includes price is used to determine the overall MAV for
each alternative as presented in Table 17.

50
Table 17.
Calculation of Overall MAV
Objectives SPIRPack SAMPack FlexSpec Packye Sentry
Normalized
weights
𝑤𝑤 𝑖𝑖
Detection rate 30 0 20 50 100 19.05
False alarm 40 35 30 25 100 17.14
Isotope ID 100 100 100 0 100 13.33
Threat
notification
mode
75 100 100 100 50 9.52
Response time 75 75 0 100 75 11.43
Weight 90 100 20 0 25 8.57
Battery 100 0 40 80 0 3.81
Post-event
support
0 100 100 50 50 7.62
Price 100 25 0 25 25 9.52
Total 610 535 410 430 525 100.00
Overall MAV 63.75 56 42.66 44 71.19

The overall MAV will be plotted against device cost in order to determine whether any
alternative is dominated and therefore excluded from consideration. Dominated devices are those
with higher costs and lower utilities. Dominated devices are those whose coordinates lie to the
right and below those of other alternatives. A circle has been drawn around the dominated
devices in Figure 4. SPIRPack and Sentry dominate the other devices. This means that
decisionmakers can reasonably remove SAMPack, FlexSpec, and PackEye from the tradespace.

51
Table 18.
Overall MAV and Cost Data
Alternatives SPIRPack SAMPack FlexSpec PackEye Sentry
Price 16.2 25 30 25 25
Weighted
utility
63.75 56 42.66 44 71.19

Figure 5.
Utility vs. Cost

The final step in our analysis is to perform sensitivity analysis of the weight assigned to
the cost attribute vs. utility. The results are presented in Figure 6. The graph confirms that
SPIRPack and Sentry dominate the other three devices. Specifically, SPIRPack should be chosen
when the decisionmaker selects a weight less than 15%; Sentry is the recommended alternative
when the weight assigned to the cost objective is greater than 15%.
0
10
20
30
40
50
60
70
80
0 5 10 15 20 25 30 35
Utility
Cost (thousands USD)
Utility vs. Cost
SPIRPack
SAMPack
FlexSpec
PackEye
Sentry

52
Figure 6.
Sensitivity Analysis – Utility vs. Cost Weight

Summary
This chapter demonstrated the application of the MAU decision methodology to the
selection of backpack radiation detection devices. The initial model was developed based upon
single-point values for each attribute; it yields a single SMART value for each device. One
shortfall of this approach is that it is based upon the input of a single stakeholder. A more
comprehensive elicitation effort would likely generate different results. Moreover, the values
assigned to detection rate and false alarm rate, arguably two of the most impactful parameters,
are notional. A model based upon actual vendor-supplied values would likely result in a different
ranking of alternatives.

53
This chapter addressed an approach for evaluating devices belonging to the same
detection archetype. When establishing a regional crisis response team, decisionmakers need to
select a portfolio of devices. Chapter 5 applies the methodology to the problem of selecting an
ensemble of devices given a set of constraints. It also discusses approaches for increasing the
“effectiveness” of a portfolio of devices based upon budget constraints, or stated differently, this
approach provides a methodology for making decisions that provide “more bang for the buck.”

54
Chapter 5: Evaluation of Radioisotope Detection Device Portfolios
Introduction
The analysis performed in this chapter broadens the application of multiattribute utility
analysis to the selection of a portfolio of detection devices. Decisionmakers may need to select a
portfolio of detection instruments that includes airborne, vehicle/vessel-mounted systems as well
as backpack, RIID, and personal radiation detections devices. They may need to decide whether
to purchase the largest number of devices that their budget will support or they may need to
select a portfolio that supports one or more operational scenarios. This chapter investigates the
following research questions:
Question 2: What criteria should be used to determine the best portfolio of devices for a
given operational scenario?
Question 3: Is it best to seek a portfolio that consists of as many devices as possible, as
long as the cost does not exceed a given price threshold?
Question 4: When selecting a device portfolio that does not exceed a given price
threshold, how do we ensure that no device in the portfolio can be replaced
by another whose performance is better?
Operational scenarios require device portfolios that support interopability, training, and
sustainability objectives. Questions 3 and 4 are not appropriate for this application because
portfolio diversification undermines operational objectives. It is not practical to deploy an
ensemble of hetereogeneous devices in operational scenarios (Questions 3 and 4).
Linear programming can be used to the optimize the selection of a device portfolio
designed by FEMA for use by first responders. This approach allows a decisionmaker to quantify

55
the cost of improving the performance of the portfolio based upon the value of an attribute or
composite measure of goodness.
Portfolio Optimization based upon Operational Scenario (Research Question 2)
Real-world operational scenarios rely upon device compatibility, consistency in
maintenance and training, and sustainability, as well as a unified approach to technology
migration—all of which are best supported by a homogeneous portfolio of devices.
This section demonstrates the use of linear programming techniques to evaluate
appropriately sized subsets of a catalog of devices detailed in Table 19. In order to simplify the
analysis, we will consider three instrument archetypes: RIIDs, backpack units, and personal
radiation detectors.
Federal and state programs provide funds to establish regional Preventative
Radiological/Nuclear Detection (PRND) Programs. FEMA provides guidance for establishing
regional PRND units. FEMA (Department of Homeland Security, 2012, 2019) provides guidance
regarding the type and number of detection units to be utilized by a PRND unit.

56
Table 19.
Device Catalog
Device Name Device Type Cost (K)/unit SMART Score
BNC NukAlert 951 PRD 1.5 43.26
BNC SAM 945 RIID 20 66.36
BNC SAM 950 RIID 22 66.4
DTect MiniRad D PRD 14 68.41
DTect MiniRad DX PRD 18 68.3
FlexSpec BRD 30 47.16
FLIR IdentiFINDER 425 RIID 2.8 64.23
Laurus RadEye PRD4 PRD 2.35 60.95
Mirion HDS 101 RIID 9.5 58.75
Mirion PDS 100 G PRD 3 66.74
Mirion PDS 100 GN ID PRD 3 77.43
Mirion SpiR-ID RIID 30 30.57
Ortec Detective 200 RIID 100 54.55
Ortec Detective X RIID 139 57.85
Ortec MicroDetective RIID 105 67.5
Ortec RADEagle RIID 16 61.93
PackEye BRD 25 41.23
Polimaster 1401 GNB PRD 4.22 47.84
Polimaster 1401MA (Gamma) PRD 1.49 47.91
Polimaster 1703MA-II PRD 1.15 67.01
Polimaster 1703MO-I PRD 1.28 73.37
Polimaster 1704A (Gamma) PRD 3 47.25
RAE Systems GammaRAE IIR PRD 1.32 75.03
RAE Systems Neutron RAE IIR PRD 2.2 69.1
SAMPack BRD 25 58.74
Sensor Tech Radiation Pager PRD 1.3 60.81
Sentry BRD 25 70.75
SPIRPack BRD 16.2 59.26
Thermo Scientific RIIDEye RIID 18 71.98

57
Figure 7 describes the personnel and equipment recommendations for a Law
Enforcement PRND unit. Our analysis does not distinguish between archetype subcategories
(Type I/II/III RIIDs or BRDs) identified in the FEMA PRND equipment configurations because
of the limited number of devices in our catalog.
Figure 7.
Law Enforcement PRND Team Requirements

We can apply MAU analysis on portfolios that could support the following scenarios:
(a) a target consisting of open areas and buildings such as an airport; (b) a fixed structure such as
a bridge or single building; or (c) natural or manmade chokepoint focused on the transport of
people or goods (Maya et al., 2011). These scenarios correspond to the overall function of the
law enforcement PRND team as defined by FEMA (Department of Homeland Security, 2019).
The equipment portfolio recommended by FEMA for the LE PRND team consists of eight to 10
PRDs, two RIIDs, and two backpack units, as indicated by the Type III equipment requirements
in Figure 7. The optimization problem is formulated as follows:

58
Optimization
• Cumulative Cost <= $250,000
• 8 ≤ Number of PRD Units ≤ 10
• Number of RIIDs = 2
• Number of BRDs = 2
The results of the optimization are detailed in Table 20.
Table 20.
Optimization Results for Law Enforcement Equipment Portfolio
Optimization Results
Actual Cost: $77,230
PRD RIID BRD
BNC NukAlert 951 FLIR IdentiFINDER 425 SPIRPack
Laurus RadEye PRD4 Ortec RADEagle SAMPack
Mirion PDS 100 GN ID

Mirion PDS 100 G
Polimaster 1401MA Gamma
Polimaster 1703MA-II
Polimaster 1703MO-I
RAE Systems GammaRAE IIR
RAE Systems Neutron RAE IIR
Sensor Tech Radiation Pager

The portfolio generated by the linear programming optimization technique satisfies the
constraints but is inappropriate for operational constraints for the reasons previously discussed.

59
The size and composition of the catalog is limited by the availability of vendor data about
candidate devices. Because vendor data were more readily available for PRDs, they comprise the
majority of the device catalog.
Table 21.
Score-Cost Ratios for Personal Radiation Detectors
Device Name Device Type Cost (K)/unit
SMART
Score
Score/Cost
Ratio
Polimaster 1703MA-II PRD 1.15 67.01 58
Polimaster 1703MO-I PRD 1.28 73.37 57
RAE Systems
GammaRAE IIR
PRD 1.32 75.03 57
Sensor Tech Radiation
Pager
PRD 1.3 60.81 47
Polimaster 1401MA
(Gamma)
PRD 1.49 47.91 32
RAE Systems Neutron
RAE IIR
PRD 2.2 69.1 31
BNC NukAlert 951 PRD 1.5 43.26 29
Laurus RadEye PRD4 PRD 2.35 60.95 26
Mirion PDS 100 GN ID PRD 3 77.43 26
Mirion PDS 100 G PRD 3 66.74 22
Polimaster 1704A
(Gamma)
PRD 3 47.25 16
Polimaster 1401 GNB PRD 4.22 47.84 11

60
Table 22.
Score-Cost Ratios for Handheld Detection Devices
Device Name
Device
Type Cost (K)/unit SMART Score
Score/Cost
Ratio
FLIR IdentiFINDER 425 RIID 2.8 64.23 23
Mirion HDS 101 RIID 9.5 58.75 6
Thermo Scientific
RIIDEye
RIID 18 71.98 4
Ortec RADEagle RIID 16 61.93 4
BNC SAM 945 RIID 20 66.36 3
BNC SAM 950 RIID 22 66.4 3
Mirion SpiR-ID RIID 30 30.57 1
Ortec MicroDetective RIID 105 67.5 1
Ortec Detective 200 RIID 100 54.55 1
Ortec Detective X RIID 139 57.85 0.42

Table 23.
Score-Cost Ratios for Backpack Detection Devices
Device Name Device Type Cost (K)/unit SMART Score
Score/Cost
Ratio
SPIRPack BRD 16.2 59.26 4
Sentry BRD 25 70.75 3
SAMPack BRD 25 58.74 2
PackEye BRD 25 41.23 2
FlexSpec BRD 30 47.16 2

61
We next consider a modification to the previous approach that results in a device
portfolio consisting of three archetypes: backpack (BRD), handheld (RIID), and body-worn
(PRD); however, the optimization only permits a single device type from each archetype in each
portfolio. Our previous example of two RIIDs, two BRDs, and eight PRDs would have resulted
in a portfolio consisting of multiple units of the same device type. The optimization approach
presented in this section selects devices in lots.
The columns corresponding to each funding level identifies the BRD, PRD, and RIIDs
that would be selected based upon the simulation. For example, the column corresponding to
40K in Table 25 indicates that a portfolio consisting of eight BNC NukAlert 951 PRDs, two
FLIR IdentiFINDER 425 RIIDs, and two SPIRPack BRDs represent the optimal mix of devices
for that funding profile. Tables 29 through 32 summarize the results of the simulations for a
range of funding levels from 25K – 300K.
Table 24.
Device Portfolios for Funding Levels 25 - 45K
Device Funding Level (K)
Name Type 25 30 35 40 45
BNC NukAlert 951 PRD
No portfolios
met this
funding
constraint
X X
FLIR IdentiFINDER 425 RIID X X X X
Mirion PDS 100 GN ID PRD X
RAE Systems
GammaRAE IIR
PRD
X
SAMPack BRD
Sentry BRD X X
SPIRPack BRD X X
Thermo Scientific
RIIDEye
RIID

62
Table 25 indicates that the same portfolio of devices is optimal for the 40 and 45K
funding levels.
Table 25.
Device Portfolios for Funding Levels 50-75K
Device Funding Level (K)
Name Type 50 60 65 70 75
BNC NukAlert 951 PRD
FLIR IdentiFINDER 425 RIID X X X X X
Mirion PDS 100 GN ID PRD X X X
RAE Systems
GammaRAE IIR
PRD X X
SAMPack BRD
Sentry BRD
SPIRPack BRD X X X X X
Thermo Scientific
RIIDEye
RIID

Table 26 indicates that the same portfolio of devices would be recommended for 50K and
60K funding levels; similarly, the same portfolio corresponds to levels 65, 70, and 75.

63
Table 26.
Device Portfolios for Funding Levels 80-100K
Device Funding Level (K)
Name Type 80 90 95 100
BNC NukAlert 951 PRD
FLIR IdentiFINDER 425 RIID
Mirion PDS 100 GN ID PRD X X
RAE Systems GammaRAE IIR PRD X X
SAMPack BRD
Sentry BRD
SPIRPack BRD X X X X
Thermo Scientific RIIDEye RIID X X X X

Table 31 indicates that the same portfolio of devices would be recommended for 80K and
90K funding levels. The same devices are recommended for levels 95 and 100. Table 32
recommends the same portfolio for funding levels between 125 and 300K. These
recommendations are limited by the size of the catalog, which only contains 29 devices, the
majority of which are PRDs.

64
Table 27.
Device Portfolios for Funding Levels 125-300K
Device Funding Level (K)
Name Type 125 150 200 300
BNC NukAlert 951 PRD
FLIR IdentiFINDER 425 RIID
Mirion PDS 100 GN ID PRD X X X X
RAE Systems GammaRAE IIR PRD
SAMPack BRD X X X X
Sentry BRD
SPIRPack BRD
Thermo Scientific RIIDEye RIID X X X X

Summary
This chapter presented approaches to selecting an ensemble of devices for use in PRND
operations. The approach showed the implementation of the FEMA’s multidisciplinary law
enforcement PRND team structure because it is flexible enough to include fire service, radiation
health, EMS, and other trained personnel, as well as sworn law enforcement personnel.
The SMART methodology was adapted to select the best-performing portfolio of devices.
It ensured that only one device model corresponding to each archetype was selected for each
portfolio. Devices would therefore be purchased in lots according to the optimization constraints.
A more thorough analysis could be performed if a richer database were available.

65
Chapter 6: Conclusion
The purpose of this professional dissertation is to provide decisionmakers with a tool that
facilitates the selection of radioisotope detection devices for preventive radiological nuclear
detection (PRND) applications. These decisionmakers could be engaged in any of the following
functions:
• PRND planning: The MAU model can support planners involved in the establishment or
enhancement of PRND operations. The model supports the selection of a single type of
device (BRD, PRD, or RIID).
• PRND training: Practitioners develop mock operational training scenarios for local, state,
tribal, and federal responders. Training activities include the selection of detection
portfolios of detection devices.
• Community policing: Local law enforcement professionals sometimes collaborate with
private, municipal, or academic partners to respond to incidents that may not necessarily
pose a threat to national security. Instead, they often involve materials such as
construction or industrial sources for which the possibility of weaponization is low.
• Domestic crisis response missions: Federal organizations such as the Department of
Energy’s Radiological Assistance Program (RAP) support the first response capability of
tribal, federal, and regional organizations. RAP’s standard equipment portfolios include
devices with a variety of detection and identification sensors.
• International crisis response missions: The Departments of Defense and State respond to
international chemical biological radiological nuclear (I-CBRN) in response to a host
nation’s request for support. Their equipment portfolios include handheld RIIDs and
backpack radiation detection devices.

66
The primary contribution of this study is the application of multiattribute utility analysis
(MAU) to the selection of radioisotope detection devices for use in detection, interdiction, and
crisis response applications. An important result is the identification of criteria for the selection
of an optimal portfolio of devices. Although diverse portfolios are preferable in some sectors,
this is not the case in PRND applications. Homogeneous device portfolios support training,
interoperability, maintainability, and sustainability mission objectives.
Assumptions and Limitations of Study
This study demonstrates the application of MAU analysis in the development of a
decision tool for use in crisis response applications. The decision model is limited in that is based
upon the input from a single PRND stakeholder rather than on the structured elicitation of
opinions from a team of subject matter experts (SME). Consequently, the methodology is applied
to notional data. The size of the device catalog was constrained by the availability of vendor
data; this limitation also affected the results.
Future Work
This project could be extended to further explore the uncertainty associated with the
distribution of attribute rankings and weights. Moreover, the analysis was performed on a
relatively small set of devices because of the lack of availability of cost and performance data. A
more robust study could be performed with access to a rich database such as the one maintained
by DHS’s System Assessment and Validation for Emergency Responders (SAVER) Program.
Moreover, the model would be improved if the attribute ranking and scoring was based upon the
input of several SMEs.
A second area of investigation could investigate the role that PRND structure has on
effectiveness. The PRND structure and organization can take a variety of forms, depending upon

67
the region’s concept of operations. The PRND can be conceptualized as a Federation-of-Systems
(FoS) because of its heterogeneity across municipal, regional, state, and federal dimensions
(Sage & Cuppan, 2001). Key characteristics of FoSs that impact the selection of equipment
portfolios are:
• Subsidiarity, which allows entities subordinate to the primary PRND team to contribute
to the decision-making process; and
• Interdependence, which allows supplemental or backup PRND units to be formed when
necessary, a characteristic that facilitates the development of a complex, adaptive
response team.
Finally, this study could be expanded to incorporate cybersecurity risks into the
evaluation criteria. These criteria would encompass issues associated with the transmission of
data from detection devices to a data fusion center. Many handheld devices and backpack
detection devices support networked reachback capability. Potential risk areas involve the
disruption, interception or spoofing of critical communications.

68
References
Baldini, E. (2012). Hunting a black swan policy options for America’s police in preventing
radiological/nuclear terrorism [Master’s thesis, Naval Post Graduate School]. Dudley
Knox Library. https://calhoun.nps.edu/handle/10945/17316
Bayes, T. (1763). An essay towards solving a problem in the doctrine of chances. Philosophical
Transactions of the Royal Society, 53, 370–418. https://doi.org/10.1098/rstl.1763.0053
Beaudouin, F. (2015). Implementing a multiple criteria model to debate about nuclear power
plants safety choices. Group Decision Negotiation, 24(6), 1035–1063.
https://doi.org/10.1007/s10726-015-9428-8
Bernoulli, D. (1954). Exposition of a new theory on the measurement of risk (English
Translation). Econometrica, 22(1), 23–36. https://doi.org/10.2307/1909829
Buddemeir, B., Musolino, S. V., & Klemic, G. (2016). Preventive rad/nuc detection equipment
categorization for consequence management. Lawrence Livermore National Laboratory.
https://doi.org/10.2172/1366955
Buddemeier, Brooke R. (2017). Mission Analysis for using Preventive Rad/Nuc Detection
Equipment for Consequence Management. United States.
https://doi.org/10.2172/1515353
Cardinal, S.-L. J., Mousseau, V., & Zheng, J. (2011). An application of constrained multicriteria
sorting to student selection. In A. Salo, J. Keisler, & A. Morton (Eds.), Portfolio decision
analysis: Improved methods for resource allocation (pp. 213–240). Springer.
https://doi.org/10.1007/978-1-4419-9943-6_10
Clemen, R. T., & Reilly, T. (2001). Making hard decisions with decisiontools. Brooks Cole.

69
Countering Weapons of Mass Destruction Act of 2018, Public Law No: 115-387 (2018,
December 21). https://www.congress.gov/115/plaws/publ387/PLAW-115publ387.pdf
de Moivre, A. (1718). The doctrine of chances. W. Pearsonn.
Department of Defense. (2016). Chemical, biological, radiological and nuclear response
publication 3-41. Department of Defense.
Department of Homeland Security. (2018a). Countering weapons of mass destruction (CWMD)
fiscal year 2019 budget justification to Congress. Department of Homeland Security.
Department of Homeland Security. (2018b). Countering weapons of mass destruction 2019
budget overview justification for Congress. Department of Homeland Security.
Department of Homeland Security. (2019, December 3). DHS Preparedness Grant Program,
Supplemental Resource: PRND guidance FY 2012.
https://www.fema.gov/pdf/government/grant/2012/fy12_hsgp_dndo.pdf
Domestic Nuclear Detection Office. (2013). Technical capability standard for backpack based
radiation detection systems. United States: Department of Homeland Security.
Edwards, W. (1971). Social utilities. The Engineering Economist summer symposium series 6
(pp. 119–129).
Edwards, W. (1977). How to use multiattribute utility for social decision making. IEEE
Transactions on Systems, Man, and Cybernetics, 7(5), 326–340.
https://doi.org/10.1109/TSMC.1977.4309720
Edwards, W., Miles, R. F., & von Winterfeldt, D. (2007). Advances in decision analysis.
Cambridge University Press.

70
Ehrgott, M., Waters, C., Kasimbeyli, R., & Ustun, O. (2009). Multiobjective programming and
multiattribute utility functions in portfolio optimization. INFOR: Information Systems
and Operational Research, 31-42. https://doi.org/10.3138/infor.47.1.31
Environmental Protection Agency. (2017). EPA radiological emergency response plan. EPA.
Fabozzi, F. J., Kolm, P. N., Pachamanova, D. A., & Focardi, S. M. (2007). Robust portfolio
optimization and management. John Wiley and Sons.
Farquhar, P. H., & Rao, V. R. (1976). A balance model for evaluating subsets of multiattributed
items. Management Science, 22(5), 528–539. https://doi.org/10.1287/mnsc.22.5.528
Fasolo, B., Morton, A., & von Winterfeldt, D. (2011). Behavioural issues in portfolio decision
analysis. In A. Salo, J. Keisler, & A. Morton (Eds.), Portfolio decision analysis (pp. 149–
168). Springer Science+Business Media.
Golabi , K., Kirkwood , C., & Sicherman, A. (1981). Selecting a portfolio of solar energy
projects using. Management Science, 27(2), 233–246.
https://doi.org/10.1287/mnsc.27.2.174
Government Accountability Office. (2018). Chemical terrorism: A strategy and implementation
plan would help DHS better manage fragmented chemical defense programs and
activities. U.S. Government Accountability Office.
IEEE Standards Association. (2013). American national standard performance criteria for
backpack-based radiation-detection systems used for homeland security. ANSI 42.53.
American National Standards Institute.
Keeney, R. L., & Raiffa, H. (1976). Decisions with multiple objectives: Preferences and value
tradeoffs. Wiley.

71
Keeney, R. L., & von Winterfeldt, D. (2011). A value model for evaluating homeland security
decisions. Risk Analysis, 31(9), 1470–1481. https://doi.org/10.1111/j.1539-
6924.2011.01597.x
Knoll, Glenn F. (2010) Radiation Detection and Measurement. Wiley
Kolm, P. N., Tutuncu, R., & Fabozzi, F. J. (2014). 60 years of portfolio optimization: practical
challenges and current rends. European Journal of Operational Research, 234(2), 356–
371. https://doi.org/10.1016/j.ejor.2013.10.060
Korhonen , P., Alexander, O., Mechitov, A., & Moshkovich, H. (1997). Choice behaviour in a
computer aided muiltiattribute decision task. Journal Multi Criteria Decision Analysis,
233–246. doi:10.1002/(SICI)1099-1360(199707)6:43.0.CO;2-S
Markowitz, H. M. (1952). Portfolio selection. Journal of Finance, 77–91.
http://links.jstor.org/sici?sici=00221082%28195203%297%3A1%3C77%3APS%3E2.0.
CO%3B2-1
Maya, I., Chatterjee, S., Salazar, D., Southwell, C., Rose, A., & Wei, D. (2011). Decision and
Risk Management Methodology for PRND Resource Allocation for the State of
California. National Center for Risk and Economic Analysis of Terrorism Events,
University of Southern California.
Mild , P., & Salo, A. (2009). Combining a multiattribute value function with an optimization
model. Decision Analysis, 139–152.
Mohres, E. (2020, March 14). Ogden City Police Department starts low-cost, high-impact
radiological safety program. Homeland Security Today. US.

72
Musolino, Stephen V.; Buddemeier, Brooke; Finfrock, Charles; Gomera, Jose; Klemic, Gladys;
Moskowitz, Joseph; Roberts, Thomas; Schaefer, Lance(2019). Evaluation of repurposing
archetypal preventive radiological/nuclear detectors for the consequence management
mission. Health Physics, 116(1), 4–17. doi:10.1097/HP.0000000000000940
National Research Council. (2013). Performance metrics for the global nuclear detection
architecture. National Academies Press.
National Security Staff, Interagency Policy Coordination Subcommittee for Preparedness and
Response to Radiological and Nuclear Threats. (2010). Planning Guidance for Response
to a Nuclear Detonation. Homeland Security Council Interagency Policy Coordination
Subcommittee for Preparedness and Response to Radiological and Nuclear Threats.
Nikou, T., & Klotz, L. (2014). Application of multi-attribute utility theor for sustainable-energy
decisions in commercial buildings: A case study. Smart and Sustainable Built
Environment, 207–222.
Office of the Assistant Secretary for Preparedness and Response. (2019, July 9). State and local
planners playbook for medical response to a nuclear detonation.
https://www.phe.gov/Preparedness/planning/playbooks/stateandlocal/nuclear/Pages/actio
nsteps.aspx
Philips, Gary W., Nagel, David J., & Coffey, Timothy. (2005). A Primer on the Detection of
Nuclear and Radiological Weapons
Pratt, J. W., Raiffa, H., & Schlaifer, R. (1965). Introduction to Statistical Decision Theory. MIT
Press.
Raiffa, H. (1968). Decision analysis: Introductory lectures on choices under certainty. Addison
Wesley.

73
Raiffa, H., & Schlaiffer, R. (1961). Applied statistical decision theory. MIT Press.
Rhodes, R. (1986). The making of the atomic bomb. Simon & Schuster.
Richardt, A., Hulseweh, B., Niemeyer, B., Sabath, F. (2010) CBRN Protection: Managing the
Threat of Chemical, Biological, Radioactive and Nuclear Weapons. Wiley-VCH.
Roosevelt, F. D. (2020, November 3). President Roosevelt's response to Dr. Einstein.
https://www.atomicarchive.com/resources/documents/beginnings/roosevelt.html
Rosoff, H. & von Winterfeldt, D. (2007) A Risk and Economic Analysis of Dirty Bomb Attacks
on the Ports of Los Angeles and Long Beach
https://doi.org/10.1111/j.1539-6924.2007.00908.x
Sage, A., & Cuppan, C. D. (2001). On the systems engineering and management of systems of
systems and federations of systems. Information Knowledge Systems Management, 325–
345. https://dl.acm.org/doi/abs/10.5555/1234195.123420
Salo, A., & Liesio, J. (2006). A case study in participatory priority-setting for a Scandinavian
research program. International Journal Information Technology Decision Making, 65–
88.
Salo, A., Keisler, J., & Morton, A. (2011). Portfolio decision analysis: Improved methods for
resource allocation. Springer Science and Business Media.
Schlaifer, R. (1951). Probability and statistics for business decisions: An introduction to
managerial economics under uncertainty. McGraw-Hill Book Co.
Sharpe, P., & Keelin, T. (1998). How SmithKline Beecham makes better resource-allocation
decisions. Harvard Business Review, 45–58.
Swan, J. S. (2004, February). A software application for multiattribute utility models. Academic
Radiology, 190–205.

74
Thrift, S. M., & von Winterfeldt, D. (2021). Risk-informed benefit-cost analysis for homeland
security R&D: Methodology and an Application to evaluating the advanced personal
protection systen for wildland firefighters. Journal of Benefit Cost Analysis, 1–32.
U. S. Army. (2009, September 1). CBRNE Consequence Management Response Force
https://www.army.mil/article/26845/cbrne_consequence_management_response_force

U. S. Department of Health and Human Services. (2019, September 30). Radiological dispersive
device playbook.
http://www.phe.gov/Preparedness/planning/playbooks/rdd/Pages/scenario.aspx
U. S. Department of Homeland Security. (2012). DHS preparedness grants. Supplemental
resource: Preventive radiological/nuclear detection guidance. DHS.
U.S. Department of Energy Energy National Nuclear Security Administration (NNSA)
Administration Office of Nuclear Incident Response. (2019). DOE radiological
assistance program – Bringing assets to bear. U.S. Department of Energy.
Von Neuman, J., & Morgenstern, O. (1944). Theory of games and economic behavior. Princeton
University Press. https://doi.org/10.1515/9781400829460
von Winterfeldt, D., & Edwards, W. (1986). Decision analysis and behavioral research.
Cambridge University Press.
Yeh, C.-H. (2003). The selection of multiattribute decision making methods for scholarship
student selection. International Journal of Selection and Assessment, 11(4), 289–296.
https://doi.org/10.1111/j.0965-075X.2003.00252.x
Yin, A. K. (1981). The evaluation of transportation systems in developing and underdeveloped
countries: The application of multiattribute utility theory and graph theories.
https://doi.org/10.25549/usctheses-c36-551209

75
Zavadskas, E., & Vaidogas, E. (2009). Multiattribute selection from alternative designs of
infrastructure components for accidental situations. Computer-Aided and Infrastructure
Engineering, 346–358. https://doi.org/10.1111/j.1467-8667.2009.00593.x

76

Abstract (if available)

Abstract The nuclear threat landscape has evolved from one dominated by the development and deployment of conventional nuclear weapons to a more likely and easily executed scenario in which an amorphous enemy launches an attack using a radiological dispersive device, more commonly referred to as a dirty bomb. The U.S. has implemented a complex, multilevel nuclear-detection architecture designed to deter, detect, and interdict attempts to transport contraband material that can be used to construct nuclear/radiological weapons of mass destruction. ❧ First responders support nuclear detection, interdiction, and crisis response missions. The technologies that support these functions can be classified into five major archetypes: body-worn, human-carried, portable, vehicle-mounted, and fixed. Collectively, these devices are referred to as radioisotope detection devices. The selection of an appropriate device or portfolio of devices involves multiple tradeoffs that are difficult for humans to perform without some kind of decision aid. This recognition provided the motivation for this dissertation. ❧ This dissertation provides decisionmakers with a tool that facilitates the selection of radioisotope detection devices for use in nuclear/radiological crisis response missions. A multiattribute utility (MAU) value model was used to identify criteria for determining the best device or device portfolio for crisis response and routine monitoring. This model can be used to guide planning, training and technology upgrade efforts implemented by local, tribal, state, federal, and international responders to nuclear crises.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A robust model for disaster risk and decision analysis

PDF

Life without nuclear power: a nuclear plant retirement formulation model and guide based on economics: San Onofre nuclear generating station case: economic impacts and reliability considerations ...

PDF

The cost of missing objectives in multiattribute decision modeling

PDF

Critical factors in evaluating compliance to United Nations Security Council Resolution 1540: developing a methodology for compliance evaluation

PDF

A strategic talent management retention model: an effective way to shape the United States Space Force

PDF

Threat assessment of physical attacks of electrical infrastructure: how do terrorist groups select their targets?

PDF

Using decision and risk analysis to assist in policy making about terrorism

PDF

Global health diplomacy: a new era of health in U.S. foreign policy

PDF

Making terrorism risk assessments more useful for decision-making

PDF

Cryptographic currency and economic security: threats, opportunities, and regulatory challenges

PDF

Gangs beyond borders: California and the fight against transnational crime - have recommendations become reality?

PDF

A time of crisis: the Australian experience and what can California learn?

PDF

Information sharing, deliberation, and collective decision-making: A computational model of collaborative governance

PDF

Lesser evils? WMD pursuit beyond nuclear weapons

PDF

A framework for evaluating urban policy and its impact on social determinants of health (SDoH)

PDF

Resilience by anticipating change: simple and robust decision making for coastal adaptation planners, communities and elected officials

PDF

Human trafficking prevention education and public school leadership: an exploratory study

PDF

Social engagement and cognitive decline in Mexican Americans: implications for age-friendly cities

PDF

A stakeholder approach to reimagining private departments of public safety: the implementation of a community advisory board recommendation report

PDF

The Human Performance Intention experiment. A quantum field framework

Asset Metadata

Creator Boadi, Antonia (author)

Core Title A multiattribute decision model for the selection of radioisotope and nuclear detection devices

School School of Policy, Planning and Development

Degree Doctor of Policy, Planning & Development

Degree Program Planning and Development,Policy

Degree Conferral Date 2021-12

Publication Date 09/22/2021

Defense Date 09/21/2021

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag counter-terrorism,homeland security,multi-attribute decision models,OAI-PMH Harvest,preventive radiological nuclear detection team,radioisotope detection

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor von Winterfeldt, Detlof (committee chair), John, Richard Sheffield (committee member), Madni, Azad (committee member)

Creator Email boadi@usc.edu,toni.boadi@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC15920947

Unique identifier UC15920947

Legacy Identifier etd-BoadiAnton-10090

Document Type Dissertation

Format application/pdf (imt)

Rights Boadi, Antonia

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu