Naval Postgraduate School
Calendar | Directory
banner
Operations Research Department Faculty Research Projects, 2010

The following are selected research projects 2013/14.

NEXT-GENERATION NETWORK SCIENCE
ResearchersDavid L. Alderson, Associate Professor
Emily M. Craparo, Assistant Professor
Thomas Otani, Professor
SponsorOffice of Naval Research
ObjectiveThe objective of this project is to conduct a broad-based, cross-disciplinary research program focused on rigorous, scalable and provably correct analysis of networks and network data. This is a Multiple University Research Initiative (MURI) Award, conducted in collaboration with colleagues at the University of Pennsylvania (UPenn), California Institute of Technology, and University of California (Santa Barbara and San Diego campuses). This was the sixth (no-cost extension) year of a five-year award.
SummaryThe Naval Postgraduate School part of the MURI team continues to focus on disaster response and management, because disasters tend to expose relationships and tensions that do not exist during everyday life. These situations are at the boundary of network science and embody two of the themes in this MURI: (1) networks with human decision-makers in-the-loop; and (2) networks that have an urgent need to take action, with lots of uncertainty. In collaboration with UCSB, we have conducted research involving behavioral social network experiments developed by the MURI team members at UPenn. This portion of the project focuses on gathering and prioritization of information from broadcast and social networks for evacuation decision-making. The first run of the experiment was deployed with civilian participants (undergraduates) at UCSB in spring 2012, with subsequent experiments at UCSB planned in the future.

 

ASSESSING RISK AND IDENTIFYING HOW TO IMPROVE RESILIENCE OF THE ENERGY SUPPLY CHAIN IN THE PACIFIC THEATRE
ResearchersDavid L. Alderson, Associate Professor
Gerald Brown, Distinguished Professor
SponsorOffice of the Assistant Secretary of Defense for Operational Energy Plans and Programs
ObjectiveThe objective of this project is to provide an integrated view of risk and resilience associated with the supply and consumption of energy to support warfighting capability in the Pacific Theater.
SummaryBuilding on several existing, but related, efforts, we focus on worst-case disruptions and their impact on operations in the United States Pacific Command (USPACOM) Area of Operations (AOR), along with prioritization, evaluation, and cost of possible mitigations. The project includes the following subtopics: (1) an analysis of the fuel distribution capability on Okinawa; (2) the optimization of prepositioned fuel farm tanks; (3) the resilience of fuel delivery via global sea routes in the presence of deliberate and non-deliberate interdiction; (4) assessing the resilience of the mainland Japan fuel distribution network; and (5) assessing the resilience of the fuel infrastructure on Oahu, Hawaii.

 

INNOVATIVE METHODS FOR ASSESSING DOMESTIC NUCLEAR SECURITY RISKS 

ResearchersDavid L. Alderson, Associate Professor
Nedialko B. Dimitrov, Assistant Professor
Kyle Y. Lin, Associate Professor
R. Kevin Wood, Distinguished Professor
SponsorDepartment of Energy, National Nuclear Security Administration (NNSA)
ObjectiveThe proposed research, funded by the Department of Energy (DOE), National Nuclear Security Administration (NNSA), as part of a larger research project on Proliferation and Terrorism Risk Assessment (PTRA), seeks (a) to develop models for evaluating adversarial risk within the transportation portion of the nuclear fuel cycle, and (b) to develop methods for making that transportation system more resilient in the face of adversarial risk. The objective of the PTRA program is to "Significantly advance the state of the art for proliferation and terrorism risk assessments."
SummaryOur study was initiated with a focus on "adversarial risk" meaning "an attack by an intelligent adversary intent on obtaining material to be used in an improvised nuclear device," but we extend the concept to include (1) direct destruction of a shipment of nuclear materials, potentially causing a release of hazardous materials, (2) the theft and transport of a shipment to a site where it is destroyed to release hazardous materials in a more damaging environment, (3) the theft of a shipment, the conversion of material into a radiological dispersion device which is later deployed as a weapon to cause more damage than would be expected from the direct destruction of a shipment, and (4) the theft of a shipment, the conversion of material into a nuclear explosive device and that device's deployment as a weapon. We focus on the United States fuel cycle, but our models are flexible enough to handle current and future fuel cycles around the world.

 

RESEARCH AND DEVELOPMENT OF TOOLS TO ASSESS THE VALUE OF ASSURED COMMUNICATIONS AND QUALITY OF SERVICES

ResearchersJeffrey A. Appleget, Senior Lecturer
Kurt Nielsen, Research Associate
SponsorLockheed Martin
ObjectiveAssist Lockheed Martin (LM) in the research and development of a wargame to investigate the effects on United States forces of attacks on satellite-provided Position Navigation and Timing (PNT) capabilities. Research updated to include meteorological and oceanographic capabilities.
SummaryThis CRADA changed from a PNT focus to examining the decaying constellation of weather satellites that Department of Defense (DoD) depends on, yet are not projected to replace. LM is developing seminar wargame events for junior and senior United States military officers, and we used the Wargaming Applications class to provide focus group feedback for LM seminar wargaming products that they developed. The student participation and feedback was very well received. In addition, Kurt Nielsen has been able to provide subject matter expertise from himself and other Department of the Navy (DoN) and DoD experts to LM, which has been very grateful for the help. This initiative will continue in FY14.

 

JWAC CHAIR AT NPS
ResearchersJeffrey A. Appleget, Senior Lecturer
Steven Hall, Research Professor
SponsorJoint Warfare Analysis Center
Objective
  • Develop and chair a mutual Naval Postgraduate School (NPS) – Joint Warfare Analysis Center (JWAC) yearly research submission/review process to screen and review potential projects for possible funding from JWAC.
  • Establish technical assessment and review process to transition prototype tools, software, and/or models to mature "cutting edge" capabilities.
  • Oversee Naval Postgraduate School (NPS) research conducted on behalf of JWAC.
  • Link up JWAC personnel with NPS subject matter experts (SMEs).
  • Bring JWAC Thesis topics to the NPS students.
SummaryA very successful site visit for Professor Alderson and Dr. Appleget was conducted in Feb 2013.
The first NPS‐JWAC research project, FOCUS (Flow of Communication upon Society) model, continued in FY13 and was successfully demonstrated in December 2013, garnering funding for FY14. Coordinated a very successful visit to NPS from senior JWAC members in August, showing them several different areas of NPS research for future leverage.

 

THE HUMAN SOCIAL CULTURAL BEHAVIOR (HSCB) MODELING INTIATIVE AT THE NAVAL POSTGRADUATE SCHOOL

ResearchersJeffrey A. Appleget, Senior Lecturer
Ronald D. Fricker, Jr., Professor
SponsorOffice of Naval Research
ObjectiveThe primary objective of the Naval Postgraduate School (NPS) Human Social Cultural Behavior (HSCB) Modeling Initiative program of research is to advance the state of the practice in Human, Social, Culture and Behavior modeling through a coherent, integrated approach leveraging basic and applied research. Two critical elements of the IW definition focus the FY 13 NPS program. The first is acquiring the understanding of how to model the relevant populations' beliefs, attitudes, intentions and behaviors. The second is modeling those indirect and asymmetric approaches designed to erode an adversary's power, influence, and will that are outlined in United States doctrine. The FY13 effort was focused on developing methods, tools, and models for use with survey data.
SummaryThis effort has developed new tools, methods, and models to summarize, analyze, and visualize complex social survey data in order to provide useful information and actionable insights for COCOMs. This research was based on six years of detailed surveys of various North African countries: 2007‐2013. The data through 2013 is maintained in-house at NPS, along with other meta‐data from the contractor who supervised the execution of the surveys.
The tasks for this effort fall into two main areas: (1) analytical methods and models, and (2) graphics and graphical displays. The former are focused on developing methods that can be used to appropriately analyze complex surveys and model the survey data in ways that provide useful and actionable information to COCOMs. The latter are focused on developing tools and techniques to graphically explore and display the data. Both areas look to address both the spatial aspects of the data, resulting from the distribution of surveys across the North African countries, and the temporal aspects of the data, resulting from the repeated application of the surveys over the past four years. We continued to recruit top quality thesis students. We graduated four thesis students this year, including one MORS-Tisdale runner up (Ben Cipperley). We conducted great outreach to SOCOM, finding with SOCOM another potential transition partner for our research. Five more thesis students have been recruited and three have competed experience tours, one to AFRICOM and two to PACOM, and two are projected for SOCOM experience tours.

 

OPTIMIZATION OF SENSOR OPERATION FOR SEARCH, SURVEILLANCE, AND RAPID ACCURATE DECISION MAKING IN MARITIME, LITTORAL AND URBAN ENVIRONMENTS

ResearchersMichael P. Atkinson, Assistant Professor
Moshe Kress, Professor
Johannes O. Royset, Associate Professor
SponsorOffice of Naval Research
ObjectiveTo develop an operational and tactical decision aid for employing sensors in an area of interest and fusing the information obtained from these sensors and from other sources.
SummaryThis project contributes to the development of a tactical decision aid (TDA) for maritime search and surveillance missions during counter-drug operations, with a particular focus on the situation faced by Joint Inter-Agency Task Force (JIATF) South. The TDA will supply the commander of counter-drug operations with a situational awareness picture, predict future locations of targets, recommend optimal allocation of search assets and effective courses of actions, and help the commander effectively utilize intelligence. The development of the TDA requires two mathematical models. A probability model takes intelligence and environmental factors as inputs and outputs a spatio-temporal probability map specifying likely locations of targets in the present and the future. The optimization model takes the probability map and operational constraints as inputs and produces a search plan for available assets.

 

OPTIMIZATION OF SENSOR OPERATION FOR SEARCH, SURVEILLANCE, AND RAPID ACCURATE DECISION MAKING IN MARITIME, LITTORAL AND URBAN ENVIRONMENTS

ResearchersMichael P. Atkinson, Assistant Professor
Moshe Kress, Professor
Johannes O. Royset, Associate Professor
SponsorOffice of Naval Research
ObjectiveTo develop an operational and tactical decision aid for employing sensors in an area of interest and fusing the information obtained from these sensors and from other sources.
SummaryThis project contributes to the development of a tactical decision aid (TDA) for maritime search and surveillance missions during counter-drug operations, with a particular focus on the situation faced by Joint Inter-Agency Task Force (JIATF) South. The TDA will supply the commander of counter-drug operations with a situational awareness picture, predict future locations of targets, recommend optimal allocation of search assets and effective courses of actions, and help the commander effectively utilize intelligence. The development of the TDA requires two mathematical models. A probability model takes intelligence and environmental factors as inputs and outputs a spatio-temporal probability map specifying likely locations of targets in the present and the future. The optimization model takes the probability map and operational constraints as inputs and produces a search plan for available assets.

 

MODELING THE DYNAMICS OF INSURGENCY AND COUNTERINSURGENCY IN A CULTURAL PRISM

ResearchersMichael P. Atkinson, Assistant Professor
Moshe Kress, Professor
Roberto Szechtman, Associate Professor
SponsorTRADOC/TRAC/TRISA
ObjectiveTo develop a model of political and insurgency dynamics in a tribal society.
SummaryThis project examines the impact of information operations (IO); how can messages affect popular opinions and behaviors. We formulate a social learning model that incorporates concepts of bias assimilation and surprising validators. Bias assimilation occurs when an individual readily accepts confirming information but rejects inconsistent evidence. A surprising validator is a trusted source who announces a stance on a topic that conflicts with the opinion of the target audience. We explore situations where the population will converge to a social opinion and others where the population polarizes into groups with extreme beliefs. We find that an effective IO campaign may require a moderate approach that appeals to individuals on extreme parts of the opinion spectrum.

 

REPLENISHMENT AT SEA PLANNER (RASP) 

ResearchersGerald G. Brown, Distinguished Professor
Walt DeGrange, Commander, Military Instructor
Anton Rowe, Research Assistant
SponsorMilitary Sealift Command
ObjectiveDevelop, install and support the Replenishment at Sea Planner (RASP), an optimization-based decision support system to aid short-term scheduling of Military Sealift Command (MSC) Combat Logistics Force (CLF) ships in a theater of operations. The goal of this scheduling is to reduce CLF fuel consumption while servicing all combatant customer ships.
SummaryIn calendar year 2013, a prototype decision support system and planner interface was developed, deployed and tested in 5th Fleet, Bahrain and in 7th Fleet in Singapore. In addition, RASP has been used to analyze the impact of withdrawal of certain MSC ships from theater. Transfer of RASP model sustainment support to the sponsor (MSC) started in Fall 2013 and will continue through FY14.

 

LARGE-SCALE OPTIMIZATION

ResearchersGerald G. Brown, Distinguished Professor
R. Kevin Wood, Distinguished Professor
SponsorAir Force Office of Scientific Research
ObjectiveThis annual proposal for continued support of our research in large-scale optimization proposes new decomposition methods for solving mixed-integer programs, attacker-defender (AD) models and defender-attacker-defender (DAD) models. AD and DAD models are sequential games that model, respectively, optimal attack and defense for military operations. These models also apply to defense of critical infrastructure in both civilian and military settings.
SummaryWe have described a three-stage sequential game (DAD) for optimal design or defense of critical infrastructure. In the simplest model, (a) a system's defender uses limited resources to protect certain system components from attack, (b) an attacker observes these defensive preparations and uses his limited resources to attack (interdict) and destroy certain unprotected components, and (c) the defender, as system operator, operates the damaged system optimally. The defender seeks a defensive plan that minimizes the worst possible damage that the attacker can inflict, when measured through the cost of post-interdiction system operations. With fixed initial defenses, DAD becomes "AD," a model for optimal system interdiction.

We have completed computation for full-scale AD and DAD analyses for the highway network in and around the San Francisco Bay Area. An extension of a traffic-equilibrium model evaluates system operations for periods of peak demand, with conditions that change over time as a function of the initial attack and subsequent repairs. A paper is under preparation.

We have submitted a paper on an AD model for analyzing a strategic spare-parts policy for an electric power grid. The second-stage of this model allocates spare parts after a coordinated attack, and then operates the system optimally over time; standard repairs are modeled deterministically. Full-scale models solve in a few hours using "global Benders decomposition." A new enumeration method solves the Benders-decomposition master problems more quickly than standard branch-and-bound, and we are extending that method for use in solving mixed-integer programs, two-stage stochastic programs and DAD models. Continuing research is extending the method to for "multi-cut formulations," which are common in stochastic programming, and to exploit parallel implementations fully.

 

LARGE-SCALE OPTIMIZATION

ResearchersGerald G. Brown, Distinguished Professor
R. Kevin Wood, Distinguished Professor
SponsorOffice of Naval Research
ObjectiveTo develop and test theory for solving large-scale optimization and game-theoretic models with applications to military planning, critical infrastructure protection, and other defense-related problems.  When appropriate, models and algorithms are implemented on request of United States combatant commanders (COCOMS), and as able, involve Ph.D. or Master's thesis research.
SummaryIn calendar year 2013, this research continued development and deployment of NMP (Navy Mission Planner) to the Navy's Command and Control Rapid Prototype Capability (C2RPC) computer system. When this ground-breaking task is complete, we will have discovered the path by which our other decision-support tools can be deployed to our fleet.

One surprise has been the quick deployment via C2RPC of a package we developed to find the shortest navigable path between any two "wet" points on the globe, a problem that turns out to be more mathematically complicated than one might expect. (A provisional United States Patent has been awarded.)
We continue theoretical development of a game-theoretic, anti-submarine model for planning defenses around a "high-value unit," typically the aircraft carrier in a carrier battle group.

We have developed and are fielding the Replenishment at Sea Planner (RASP), with initial installation completed at 5-th Fleet, Bahrain, and 7-th Fleet, Singapore. RASP minimizes fuel consumption by Navy supply ships as they attend deployed combatants and has been certified by independent auditors to save a lot of fuel (U).
Our transit fuel planner has been deployed on Littoral Combat Ships LCS-1, 2 and 3, and can save as much as 20% fuel consumption. This year the planner also deployed with CG 65 Chosin in the western Pacific.

We continue to develop and support the Combat Logistics Force (CLF) strategic and operational planning models, and have evaluated current war plans for the 7-th Fleet, among others, for logistic feasibility.

We continue to apply our missile-defense model, Joint Defender (JDEF) for EUCOM to reckon where to locate, and re-locate defensive radars and interceptor missile fields in Eastern Europe. Changes in political policy and diplomacy have necessitated plan revisions.

The distinguishing ability we bring to bear here is complete mathematical mastery and ownership of these optimization-based decision-support tools, combined with deep domain expertise in warfare modeling, with reach-back support offered via a single phone call or email at any classification level. Some requests require exigent mathematical developments, and some just seasoned modeling advice.

 

NAVY RECRUIT ATTRITION PREDICTION MODELING

ResearchersSamuel E. Buttrey, Associate Professor
Lyn R. Whitaker, Associate Professor
Ronald D. Fricker, Jr., Professor
SponsorOPNAV N1 and Commander, Naval Recruiting Command
ObjectiveUsing information from the Active Duty (AD) Master File and the Military Entrance Processing Command (MEPCOM) file, this project will develop a model that most precisely predicts a sailor's likelihood or probability of surviving (i.e., not attriting) through the 1st term of enlistment. It will then compare and contrast the predictive performance of the model versus the current Quality Matrix with the goal of judging whether the model performs significantly better than the Quality Matrix. The goal of this task is to gain some insight into whether improvement is possible and, if so, what the magnitude of the improvement is likely to be.
SummaryThe research only started in late 2013 and will be on-going through later 2014.

 

ANALYSIS SUPPORT FOR COMPREHENSIVE SOLDIER FITNESS 

ResearchersSamuel Buttrey, Associate Professor
SponsorTRAC-Monterey
ObjectiveExamine responses to the GAT survey with emphasis on the effect of resilience trainers
SummaryThe Global Assessment Tool (GAT) is a survey administered to every soldier at least once a year. The survey measures the soldier's fitness in emotional, family, social and spiritual domains. An important part of the Army's Comprehensive Soldier Fitness (CSF) strategy is to train and embed mater resilience trainers (MRTs) in as many units as possible. The average effect associated with a soldier being in a unit with an embedded MRT can be measured, albeit imperfectly, by examining GAT scores in units with and without MRTs. This research showed a small but non-zero effect of the presence of MRTs, and in particular observed that the different locations at which MRTs were trained produced essentially the same-sized effects.

 

ARMY STUDY TO ASSESS RISK AND RESILIENCY IN SOLDIERS (STARRS) VALIDATION AND ANALYSIS 

ResearchersSamuel Buttrey, Associate Professor
SponsorTRAC-Monterey
ObjectiveDuplicate the statistical analysis performed by the STARRS research team.
SummaryThe Army Study to Assess Risk and Resilience in Service members (Army STARRS) is, according to the STARRS web page, the largest study of mental health risk and resilience ever conducted among military personnel. In this project we set out to reproduce and validate the statistical analysis and results from Army STARRS based on syntax and variable coding provided by the original STARRS research team. This project also served as the first full-scale exercise of the Person-Data Environment, an approach that gives researchers access to manpower data in a "sealed" virtual desktop so as to protect the confidentiality of the data.

 

MILITARY APPLICATIONS OF OPTIMIZATION 

ResearchersW. Matthew Carlyle, Professor
David L. Alderson, Associate Professor
Emily M. Craparo, Assistant Professor
Javier Salmerón, Associate Professor
SponsorOffice of Naval Research
ObjectiveThe proposed research was to apply appropriate optimization models to relevant decision problems in the Navy and, more broadly, in the Department of Defense (DoD).
SummaryThis research involves the rapid development and deployment of medium- to large-scale optimization models and prototype Microsoft Excel/VBA decision support tools that assist decision making at the strategic and operational levels. These tools are either transitioned directly to DoD analysts who use them, or allow us to provide reachback analysis at all levels of classification. In support of this effort we are also developing a computational infrastructure to provide web-based access to our decision support tools that will simplify the transition process.

 

AERIAL RECONNAISSANCE & SURVEILLANCE OPTIMIZATION MODEL DEVELOPMENT

ResearchersEmily M. Craparo, Assistant Professor
SponsorTRAC-Monterey
ObjectiveIn an effort to preserve the Army's unmatched capabilities in aerial reconnaissance and surveillance (R&S), the Integrated Capabilities Development Team (ICDT) administered a large-scale study during fiscal years 2012 and 2013 to determine in which R&S platforms and sensors the Army should invest. Part of this effort was the formulation and implementation of the Joint Platform Allocation Tool (JPAT), an integer linear program designed to determine an optimal R&S investment portfolio by evaluating cost, performance, and production timelines of existing and planned assets, as well as these assets' ability to perform against a 12-year prioritized mission demand signal.
SummaryThe JPAT model was refined in conjunction with collaborators at TRAC and numerous subject matter experts. Refinements included development of a rolling horizon implementation, addition of new constraints, and incorporation of additional preprocessing and post-processing capabilities. JPAT has informed critical resourcing decisions concerning the Army's long-term investment strategy.

 

NUCLEAR PROTOTYPE TRAINING ALLOCATION 

ResearchersRobert F. Dell, Professor
SponsorDepartment of Energy, Knolls Atomic Power Laboratory
ObjectiveThe goal of this research is to develop and implement an integer linear program to prescribe how many students of each class and type to allocate to each Nuclear Power Training Unit site.
SummaryA prescriptive optimization model has been developed and implemented. The model has been adopted by the sponsor and has completely replaced the sponsor's legacy program.

 

PLANNING ENVIRONMENTAL CLEANUP AT CLOSED ARMY INSTALLATIONS 

ResearchersRobert F. Dell, Professor
SponsorOffice of the Army Assistant Chief of Staff for Installation Management
ObjectiveThe goal of this research is to develop and implement of an integer linear program to prescribe how to allocate funding for environmental cleanup at closing Army installations.
SummaryThe investigator is providing research, support, and development of optimization models to assist the Army's Assistant Chief of Staff for Installation Management.  Updating the integer-linear program BAEC (Budget Allocation for Environmental Cleanup) was the primary 2013 development effort.  In 2013, the Army used BAEC to help plan over $500 million in environmental cleanup. 

 

APPLICATIONS FOR THE BIOSENSE MONITORING TOOL

ResearchersNedialko B. Dimitrov, Assistant Professor
SponsorTexas Department of State Health Services via CRADA with UT Austin
ObjectiveThe aim of this project is to determine effective uses of BioSense 2.0 data by DSHS central, regional and local public health preparedness programs for detecting early stage outbreaks of infectious disease threats, and monitoring transmission, morbidity and mortality of such outbreaks, particularly for at-risk populations. The aim is additionally to use historic data to quantitatively assess the performance of BioSense 2.0 in influenza surveillance systems, especially in its ability to provide information on Texas' at-risk populations.
SummaryIn 2013, work on the research project concluded. In collaboration with the University of Texas at Austin, we prepared a report on the performance of BioSense 2.0 data in influenza surveillance. Initial attempts to access BioSense 2.0 data directly were unfruitful because of data use agreements between parties in Texas. Instead, we used Google's public time series of BioSense 2.0 emergency department visits in the Dallas area and surrounding counties for the assessment. The analysis shows that BioSense 2.0 does indeed provide significant improvement on top of the traditional ILINet surveillance system. However, the improvement provided by BioSense 2.0 decreased when Google Flu Trends was also included. The analysis was based on a novel method of predicting influenza hospitalizations using a Poisson processes model. We aim to submit a publication on the results to a scientific journal in 2014.

 

EFFICIENT INTELLIGENCE PROCESSING FROM LARGE NETWORKS

ResearchersNedialko B. Dimitrov, Assistant Professor
Moshe Kress, Professor
SponsorUSMC
ObjectiveThe aim of this project is to develop mathematical models and methods to assist intelligence processors on identifying relevant data in a glut of information. Specifically, the project concerns two separate scenarios. In the first scenario, intelligence processors have a collection of person-to-person communications to screen. The processors do not have sufficient time to screen all the communications, and the mathematical model is to help identify communications that are likely to be relevant. In the second scenario, intelligence processors have a collection of documents. Similarly to the first scenario, the processors do not have enough time to screen all the documents. The mathematical models developed will assist in focusing the processor's attention on documents that are likely to be relevant.
SummaryIn 2013, work on the research project began. A master's student, Duncan Ellis, developed a modular software system on which to test both optimization and inference algorithms for the person-to-person communication scenario. We used this software system to run experiments using the Enron corpus of emails. In addition, we recruited a second master's student, Chris Wood, to assist in developing models and software for the second scenario, identifying relevant documents. We initiated interactions with several researchers in other universities, including Oren Kurland at the Technion and Kevin Glazebrook and Paul Fearnhead at Lancaster University. Finally, we recruited a PhD student at Lancaster University, Lisa Turner, to assist with the modeling effort. We have completed initial mathematical modeling and computational experiments for the person-to-person scenario, and have submitted a manuscript for review.

 

NETWORK RECONFIGURABILITY: DESIGN, SENSING, CONTAINMENT, AND RECOVERY TO WEAPONS OF MASS DESTRUCTION ATTACK

ResearchersNedialko B. Dimitrov, Assistant Professor
SponsorDefense Threat Reduction Agency via CRADA with UT Austin
ObjectiveCollaborators will develop tools, theory, and methodology, for a new dynamic, modular, and reconfigurable network theory, geared to enable resilience, resistance, and recovery of our network infrastructure to Weapons of Mass Destruction (WMD) attack. Thwarting potential attacks, and containing and rapidly recovering from a successful attack, requires the efficient exploitation of the vast existing military and civilian interconnecting infrastructure, and cyber-infrastructure networks. WMD-stressors present new challenges to network design and related recovery efforts that require new analysis and Synthesis techniques, new estimation techniques, algorithms, and information processing.
SummaryIn 2013, work on the research project concluded. A publication on the main results, discussing new algorithms to solve budget-constrained Markov Decision Processes (MDP) was submitted to IEEE Transactions on Automatic Control. In multiple edits on the manuscript, the theoretical guarantees provided by the algorithms were outlined in detail. The algorithms provide running times much faster than the other available solution method, linear programming. The new algorithms are faster by a multiplicative factor equal to the number of states in the MDP, creating a significant contribution to the literature of constrained MDPs.

 

RIGHT SIZE VIROLOGIC SURVEILLANCE PROJECT

ResearchersNedialko B. Dimitrov, Assistant Professor
SponsorAssociation of Public Health Laboratories via CRADA with UT Austin
ObjectiveThe aim of this project is to introduce statistical guarantees into the U.S. National Influenza Surveillance system. In particular, the project concerns the methods and quantities with which public health laboratories sample influenza virus from the population. The project's goal is to develop a set of right-size calculators, along with documents for guidance to help public health lab operators understand how they can contribute to the national influenza surveillance efforts. In conjunction with the APHL and CDC, the right-size calculators will help the independently operated laboratories cooperate as a system to provide statistical guarantees nation-wide.
SummaryIn 2013, work on the research project concluded. Based on data provided by the APHL and CDC, we analyzed past sampling efforts, and constructed methodology for introducing statistical guarantees into the influenza virus sampling process. A set of four right-size calculators were developed, along with accompanying documentation, user guides, and tutorials. The resulting right-size calculators were demonstrated to over 82 nation-wide public health laboratories in August of 2013, and are currently in use to help guide the U.S. Influenza surveillance efforts. In addition, a preliminary analysis was performed on the biases present in the surveillance system.

 

OPNAV CHAIR OF SYSTEMS ENGINEERING ANALYSIS 

ResearchersJames N. Eagle, Professor
SponsorOPNAV N9I
ObjectiveProvide support and liaison roles to OPNAV N9I in the administration of the Systems Engineering Analysis curriculum.
SummaryDuring calendar year 2013, the OPNAV Systems Engineering Analysis (SEA) chair completed the following:
  • Supported N9I in its role as the sponsor of the SEA curriculum.
  • Identified opportunities for Naval Postgraduate School faculty and students to support N9I in carrying out studies and analyses and coordinating resulting cooperative efforts.
  • Participated in research and instructional activities in support of the SEA and related curricula.
  • Cooperated with the Operations Research and Systems Engineering departments in administration of the SEA program.
  • Provided for the overall coordination of two large, interdisciplinary, capstone projects as part of SEA degree requirements, and ensured N9I concurrence with the content and aims of such projects.

 

FDA ANALYSIS SUPPORT 

ResearchersP. Lee Ewing, Research Associate Professor
SponsorArmy G-8, Forced Development Directorate, Warfare Analysis Division
Objective

The research team will continue to provide study, analysis, modeling and simulation, and use of analytical tools support to assist the Chief of the Army G-8 Force Development Directorate's Warfighting Analysis Division (DAPR-FDA) in the fulfillment of the division's mission. The study focuses on:

  • Providing DAPR-FDA study, analytical, modeling, and technical support to develop a discrete event simulation model (DES) which will replicate established Army equipping processing while incorporating primary uncertainties that may affect senior leader equipping decisions.
  • Once a prototype discrete event simulation model is developed, investigate optimizing the simulation or develop a separate optimization model to provide optimal policies as inputs for the discrete event simulation model.
  • Support a student experience tour between the Naval Postgraduate School (NPS) Operations Research (OR) department and DAPR-FDA to facilitate the continued development of DAPR-FDA models and analysis techniques.
SummaryThe objective of this work changed from development of a discrete event simulation model to further development and extending the capabilities of the existing optimization models used by the sponsor. Work continues on this project into 2014.

 

CPAT TECHNICAL VERIFICATION

ResearchersP. Lee Ewing, Research Associate Professor
SponsorProgram Executive Office for Ground Combat Systems
ObjectiveThe principle investigator provides a technical verification and validation of the Capability Portfolio Analysis Tool (CPAT).
SummaryThis project consists of one primary deliverable, production of a final Naval Postgraduate School technical report. The report addresses the following: i) Document and interview methodology review. ii) Verification of the qualitative model structure, and quantitative value functions and swing weights. iii) Verification of the optimization formulation and OPL code. iv) Validation of the decision analysis model and swing weight sensitivity analysis. v) Validation of the optimization model.

 

OSD CAMPAIGN ANALYSIS 

ResearchersP. Lee Ewing, Research Associate Professor
SponsorOffice of the Secretary of Defense, CAPE
Objective

The research team will study and suggest alternative approaches to traditional large-scale campaign simulation models by demonstrating the utility of Scenario-Based Warfighting Analysis (SBWA) with traditional Operations Research tools and methods and simple campaign modeling tools such as the Fast Theater Model (FATHM). In general, the study's primary tasks are to:

  • Develop an agile campaign analysis approach which emphasizes resource relationships between two or more of the Services and allows for rapid development and modeling of alternative CONOPs and scenarios.
  • Demonstrate the approach by providing analytical products based on two scenarios and CONOPs.
  • Summarize the campaign analysis approaches used and study results in a written report. Include strengths and weaknesses of the approaches used and any dependencies on traditional large scale campaign simulation models
SummaryThis work is conducted by two Naval Postgraduate School teams. One team consisting of Dr. Brown and Dr. Washburn, on contract from Institute for Defense Analysis (IDA), has modified FATHM to demonstrate its use for campaign analysis. The other team, led by Mr. Kline, has produced a CAPE approved scenario to demonstrate Scenario-Based Warfighting Analysis. Mr. Kline and his team are using traditional Operations Research tools to conduct this analysis. Both efforts are ongoing and are scheduled to complete in FY14.

 

FUTURE NAVY RECRUITING STRATEGIES: DEVELOPMENT OF CONCEPTS, PRACTICES, PROCESSES AND PROCEDURES TO RECRUIT GENERATION Z AND ALPHA 

ResearchersRonald D. Fricker, Jr., Professor
Samuel E. Buttrey, Associate Professor
LTC Jonathan K. Alt, Military Assistant Professor
SponsorOPNAV N1 and Commander, Naval Recruiting Command
ObjectiveThis research project will identify and evaluate past alternative recruiting efforts with an emphasis on assessing the quantitative evidence (if any) of performance. It will also assess the literature on trends in the Millennial and post-Millennial generations with a focus on how well current and previous recruiting efforts align with how these generations are likely to want to interact with Navy recruiting. Finally, this research will pose possible future recruiting strategies for the Navy along with rigorous quantitative methods for evaluating the performance of the various strategies and their components.
SummaryThe research started in late 2013 and will be on-going through later 2014.  The 2013 effort focused on conducting an open literature review in order to summarize:
  • Navy recruiting processes and infrastructure;
  • Characteristics and behaviors of the millennial (and later) generations that are relevant to military recruiting;
  • Ideas and proposals for improving the efficiency/effectiveness of military recruiting; and,
  • Past military recruiting experiments and their results.

 

ANALYSIS OF CAPABILITY PERFORMANCE REVIEWS 

ResearchersP.A. Jacobs, Distinguished Professor
SponsorTRADOC Analysis Center (TRAC)-Monterey
ObjectiveThe Secretary of the Army directed Capability Portfolio Reviews (CPR) as a pilot process to holistically examine the requirements that drive capability development, acquisition and sustainment to determine if current and proposed programs are aligned to meet key national and defense strategies and Army plans. A CPR includes the scoring on a categorical scale by subject matter experts (SMEs) of the contribution of future capabilities to the successful conduct of military tasks. Statistical techniques including Bayesian methods are explored to summarize such data to assist decision-makers in resource allocation decisions.
SummaryGraphical analysis of simulated data from TRAC is conducted. A Bayesian approach to analysis of CPR data is proposed.

 

DESIGN OF EXPERIMENTS FOR FOLLOW-ON OPERATIONAL TEST OF THE AEGIS MODERNIZATION PROGRAM 

ResearchersP.A. Jacobs, Distinguished Professor
SponsorNaval Sea Systems Command
ObjectiveSupport and strengthen the Aegis Modernization (AMOD) program by guiding and assisting in Developmental Tests (DT), Integrated Tests (IT), Initial Operational Test and Evaluation (IOT&E), Follow-On Test and Evaluation (FOT&E) planning. Advise on use of Design of Experiments (DOE) methods to AMOD T&E. Carry out test preview using analytical-mathematical-probabilistic models. Such steps will expedite choice of test design factors (e.g. attacker/raid type, speed, altitude, course, etc.) and quantitative factor levels.
SummaryModels for quantifying simulation model prediction accuracy using at-sea test results have been formulated and studied.

 

UNDERSTANDING OPTIMAL DECISION MAKING IN WARGAMING 

ResearchersQuinn Kennedy, Senior Lecturer
LTC Jonathan K. Alt, Military Assistant Professor
Ronald D. Fricker, Jr., Professor
Jeffrey A. Appleget, Senior Lecturer
SponsorArmy Research Office
ObjectiveTo objectively and quantitatively define and measure optimal decision making in wargaming through the use real time measurements of visual scan patterns and brain activity.
SummaryThe work described in this section was conducted in year 1 of a three-year effort. The researchers are investigating optimal wargaming decision making with a multipronged approach across two studies. Study 1 focuses on the development of optimal decision making when all subjects begin as naïve decision makers. Specifically, the researchers attempt to identify the transition from exploring the environment as a naive decision maker to exploiting the environment as an experienced decision maker via statistical and neurological measures. Two cognitive abilities required for optimal decision making in wargaming are reinforcement learning, the ability to learn from trial and error; and cognitive flexibility, the ability to recognize when the rules have changed or that the current strategy no longer works. Therefore, the researchers developed two wargames (tasks) that tap these abilities. While volunteer warfighters complete each wargame, their decisions, visual scan patterns and brain activity are tracked in real time. Study 2, a Naval Postgraduate School thesis study examines wargaming decision making in a dynamic and complex environment. Volunteer warfighters will be placed in a virtual environment created via Virtual Battlefield Simulation 2 (VBS2) in which they are the leader of a Bradley vehicle section and periodically, must make clearly defined tactical decisions.

 

DEVELOPMENT OF OPERATIONAL PLANNING TOOLS FOR THE MARITIME OPERATIONS CENTER 

ResearchersJeffrey E. Kline, Professor of Practice
Gerald G. Brown, Distinguished Professor
R. Kevin Wood, Distinguished Professor
W. Matthew Carlyle, Professor
Javier Salmerón, Associate Professor
Anton Rowe, Research Associate
Carol O'Neal, Research Associate
SponsorOffice of Naval Research
ObjectiveDevelop and evaluate decision aids for use by Navy staffs to plan maritime operations.
SummaryThis research will develop, produce, evaluate and deliver complete, operational prototypes of various, optimization-based planning systems for supporting the maritime operations center and maritime headquarters planning staff in maritime operational missions including: strike; information, surveillance, and reconnaissance; theater security and cooperation; theater ballistic missile defense; anti-submarine warfare; logistics routing; transit planning; maritime interdiction operations; and others. In 2013 this program further advanced prototype decision aids related to maritime operational level planning. The ship transit planner, a U.S. Navy patented optimizing tool that was developed under this program was modified for use by the USS CHOSIN (CG-65). The Oceanic Routing Service (ORS), U.S. Navy patent pending, is a program to create shortest-path transits that avoid obstacles between two oceanic locations, was modified into a web-based "widget" and integrated into the PEO C4I's Command and Control Rapid Prototype Capability (C2RPC), were it is said to be the most-used utility the system offers. Work advances on providing the Navy Mission Planner (NMP), renamed the Mission Planning Service, as a tool for use in C2RPC. This research program also provided student support in developing the Replenishment At Sea Planner (RASP) and thesis travel in critical infrastructure defense.

 

HIGH ENERGY LASER EMPLOYMENT IN SELF DEFENSE TACTICS ON NAVAL PLATFORMS 

ResearchersJeffrey E. Kline, Professor of Practice
Thomas W. Lucas, Professor
SponsorAssistant Secretary of the Navy (Research and Development)
ObjectiveDevelop an agent-based model to assess the high energy laser's contribution to ship self-defense.
SummaryThis project proposes to model various tactical situations to evaluate the HEL's impact on ship self-defense in high threat environments using simulation tools like Map Aware Non-uniform Automata (MANA), SIMKIT, ARENA, HELEEOS, EADSIM and other tailored models from NSWC Dahlgren W10 Warfare Analysis division. The Monte Carlo simulation will be used to explore the performance of the DDG combat system equipped with a HEL or conventional KE weapons. Leveraging off modeling efforts at NSWCDD and past studies conducted at NPS, a DDG in a demanding ship's self-defense tactical situation will be simulated using a NSWCDD engagement model and by creating a Naval Postgraduate School agent-based model in MANA. The results from both efforts will be individually analyzed and then compared and contrasted. This is currently a program in progress with an initial MANA model or "SSLS in MANA" created.

 

NAVY WARFARE DEVELOPMENT COMMAND'S OPERATIONS RESEARCH CHAIR OF WARFARE INNOVATION 

ResearchersJeffrey E. Kline, Professor of Practice
Carol O'Neal, Research Associate
Lyla Englehorn, Research Associate
SponsorNavy Warfare Development Command
ObjectiveThis long standing Chair is based most recently on Memorandum of Agreement between Naval Postgraduate School and NWDC designating the Chair's purpose to "…invigorate and conduct research and analysis required to develop doctrine, tactics, techniques, procedures, and maritime and joint operational concepts" at the Naval Postgraduate School.
SummaryThe Chair sponsors faculty and student research in a variety of areas. During calendar year 2013 this included several warfare innovation workshops dealing with use of unmanned systems and advanced undersea warfare systems; special mini-studies within the Joint Campaign Analysis and the JC4I capstone classes; Project Jason, a research series to counter UAVs; continued development of maritime operational planning aids; and a variety of student travel in support of field experimentation and research trips.

 

STATISTICAL RESEARCH TO SUPPORT AEGIS SPY-1 OPERATIONS 

ResearchersRobert A. Koyak, Associate Professor
Rachel T. Silvestrini, Assistant Professor
SponsorU.S. Navy Program Executive Office, Integrated Warfare Systems
ObjectiveThe project focuses on two areas where statistical research can lend insight to improve SPY-1D(V) radar performance. The first is development of prognostics algorithms in order to predict remaining life (RUL) or to give an indication that maintenance of a SPY-1 subsystem is required in a near time frame. The second is an evaluation of alternative business rules for managing carcasses of repairable SPY-1 inventory items. Both areas are addressed in separate research tasks.
SummaryThis project examined data from several sources to determine their value in the development of condition-based maintenance procedures for SPY-1D(V) radar systems. It also developed a simulation model for the movement of 10 KW traveling wave tube (TWT) carcasses used in SPY-1 radars in order to find opportunities to improve the efficiency of throughput.

 

IMPROVED FORECASTING METHODS FOR NAVAL MANPOWER STUDIES 

ResearchersRobert A. Koyak, Associate Professor
Lyn R. Whitaker, Associate Professor
SponsorOPNAV N15
ObjectiveThe purpose of this project is to conduct an investigation into the statistical properties of recently-developed forecasting methods, and other emerging methods that may be identified as having desirable attributes, using data provided by NPRST that encompass recent events in the U.S. economy (e.g. the 2008 housing market debacle), and changes that affect the mission and budget of the Navy.
SummaryThis project is part of the Naval Studies Program. Initial funding was received in late November 2013. Initial discussions with NPRST took place in calendar year (CY) 2013 to provide scope for the project as it continues into CY 2014. The Master's thesis research of LCDR Sean McCrink (USN) on modeling promotion probabilities for enlisted Sailors also is being sponsored under this project. In December 2013 LCDR McCrink visited NPRST in Millington, TN to obtain information that will be useful to his research. He is expected to complete his thesis by June 2014.

 

FA-XX LONG-TERM COST AND TECHNOLOGY RISK 

ResearchersRobert A. Koyak, Associate Professor
Gerald G. Brown, Distinguished Professor
SponsorOPNAV N98G
ObjectiveThis project develops an application of the approach used in Brown, Grose, and Koyak (2006) to FA-XX or other components of the Navy's Thirty Year Aviation Plan. Depending on the availability of suitable data during FY 2014, development of the approach on a smaller-scale application may be considered in consultation with the sponsor. Upon finalization of the optimization model it will be simulated in order to exercise its stochastic elements. The number of simulations will be chosen so that attributes of cost and timeliness are estimated with acceptable accuracy, and to produce prediction intervals of desired quantities. A software version of the optimization model will be developed using Visual Basic for execution in Microsoft Excel.
SummaryThis project is part of the Naval Studies Program. Initial funding was received in late November 2013. Initial discussions with N98G took place in calendar year (CY) 2013 to provide scope for the project as it continues into CY 2014.

 

WHAT ARE THE OPTIMAL MAINTENANCE INTERVALS FOR PMCS ON MARINE CORPS EQUIPMENT? 

ResearchersRobert A. Koyak, Associate Professor
Michael P. Atkinson, Assistant Professor
SponsorMarine Corps Logistics Command
ObjectiveThis project investigates development of preventative maintenance schedules for U.S. Marine Corps (USMC) land vehicles based on condition and usage information. Current schedules rely on chronological time, which is an inadequate substitute for condition and usage. Information to be considered includes, but is not limited to: mileage, cumulative engine cycles, operating environment (e.g. climate, terrain), and past repair history. The class of vehicles to which this effort will be applied includes MTVRs and MRAPs, particularly those equipped with engines (e.g. Caterpillar) that have automatic data-capture capabilities.
SummaryThis project is part of the Naval Studies Program. Initial funding was received in late November 2013. Initial discussions with MARLOGCOM and its contractors took place in calendar year (CY) 2013 to provide scope for the project as it continues into CY 2014.

 

MISSION COMMAND ANALYSIS USING MONTE CARLO TREE SEARCH

ResearchersKyle Y. Lin, Associate Professor
SponsorTRAC-Monterey
ObjectiveTRAC-Monterey is carrying out research to apply and evaluate Monte Carlo Tree Search (MTCS) method to tactical decision situations in a simulated environment in order to expand mission command analysis capabilities. The objective of this project is to support this effort by developing efficient algorithms to benchmark and improve the MCTS method.
SummaryTRAC-Monterey is carrying out supporting research to produce a documented and tested methodology that applies Monte Carlo Tree Search (MCTS) methods to decision situations in order to expand mission command oriented analysis. Mission command features decentralized execution with subordinate commanders exercising disciplined initiative while acting aggressively and independently to accomplish the mission within the commander's intent. The major task of this project is to develop efficient algorithms for a near-optimal fire allocation policy that can be used to benchmark and improve the MCTS method.

 

OPTIMAL SURVEILLANCE PATROL WITH EXTENSIONS

ResearchersKyle Y. Lin, Associate Professor
Michael P. Atkinson, Assistant Professor
Timothy Chung, Assistant Professor
SponsorOffice of Naval Research
ObjectiveThis research proposes to continue to study a class of problems in which a defender dynamically allocates its surveillance assets in anticipation of an attack. The surveillance asset can be an unmanned vehicle, a camera mounted on a blimp, or a patrol officer. The proposed research will develop robust methods to counter the enemy's strategy in the worst-case scenario
SummaryThis project extends earlier work in two directions. First, the new model accounts for the travel distance between search areas, and the time it takes to search an area. On this direction, the research team will formulate an optimization model to compute the optimal solution, and develop efficient heuristic methods that are expected to deliver near-optimal results. Second, the research team will study how to carry out patrol plans when there are multiple patrollers, in order to minimize expected damage caused by attacks from an intelligent and adaptive adversary.

 

FACTOR SCREENING IN A HIGH-DIMENSIONAL SYSTEM OF SYSTEM SIMULATION

ResearchersThomas W. Lucas, Professor
SponsorArmy Research Laboratory
ObjectiveEmbed the System of Systems Survivability Simulation (S4) in a data farming environment and test this new capability with an assessment of S4's treatment of communication.
SummaryThe Survivability/Lethality Analysis Directorate (SLAD) of the U.S. Army Research laboratory (ARL) analyzes the survivability, lethality, and vulnerability (SLV) of current and future battlefield systems. To better account for the total force effects of information flows over a common network, SLAD, in collaboration with the Physical Science Laboratory (PSL) at New Mexico State University (NMSU), is developing a System of System (SoS) simulation. To reach its full potential, analysts must be able to efficiently determine the subset of input factors that exert the greatest impact on key response variables. To enable this capability, the Naval Postgraduate School's (NPS's) SEED Center for Data Farming partnered with ARL, SLAD and NMSU PSL in developing appropriate factor screening methods for multiple outputs. Information transfer included a short course in which at least one new algorithm was presented.

 

DEVELOPING SYNTHETIC THEATER OPERATIONS RESEARCH MODEL (STORM) ANALYTIC UTILITY 

ResearchersThomas W. Lucas, Professor
SponsorOPNAV N81
ObjectiveTo develop tools and processes that reduce the amount of manpower and time required to complete STORM output post-processing.
SummaryA modeling environment that underpins many important N81 and joint studies is the Synthetic Theater Operations Research Model (STORM). The Navy and other Services use STORM as a tool to evaluate campaign risk and assess the utility of operational and acquisition decisions. STORM is a large, stochastic, campaign-level simulation that requires extensive detail for both system parameters and operational concepts and generates a huge amount of output data. Because STORM is stochastic, multiple replications are made for given inputs. A current impediment to fast and efficient use of STORM is the volume of data it generates. The objective of this study is to develop tools and processes to reduce the amount of manpower and time required to complete STORM output post-processing. This will enable a greater number of dynamically determined scenario replications and improve the speed and accuracy with which analysts will be able to gather insights.

 

DEVELOPMENT OF A PROTOTYPE IMPROVED PERFORMANCE ASSESSMENT AND READINESS TRAINING SYSTEM 

ResearchersMichael E. McCauley, Research Professor
Mathias Kolsch, Associate Professor
SponsorNAVAIR PMA 205 (NAVAIR Training)
ObjectiveContinue the development and testing of the working prototype Landing Signal Officer (LSO) Improved Performance and Readiness System (IPARTS).
SummaryLimited funding was available to move this system forward in the Navy acquisition system, despite strong pleas from the LSO community to make this system available for use in the fleet. IPARTS replaces the legacy system (APARTS) to enable LSOs to capture and store carrier landing outcomes in a database for rapid retrieval and performance feedback to pilots. This system was identified by the sponsor, Office of Naval Research (ONR) TechSolutions last year as one of their top three "success stories." However, there is a substantial gap between a successful prototype and getting it ready and approved for implementation on United States aircraft carriers. We are attempting to bridge that gap for IPARTS.

 

PERFORMANCE SHAPING FUNCTIONS OF ENVIRONMENTAL STRESSORS: VIBRATION AND SHIP MOTION 

ResearchersMichael E. McCauley, Research Professor
SponsorOffice of Naval Research Code 34
ObjectiveIdentify data and research findings relevant to the development of models of the influence of environmental stressors on human performance in Navy operational settings. Investigate existing guidelines for the magnitude of lateral acceleration that causes postural instability (a "motion induced interruption"). Participate in at-sea data collection LCS and other vessels and coordinate with other Navy laboratories on data analysis and report writing.
SummaryShip motion can affect human performance in several ways, depending on variables such as sea state, vessel heading relative to the seas, vessel speed, hull form, location on the vessel, and the type of task being performed. This Office of Naval Research (ONR) project coordinated a set of Navy laboratories in addressing these issues on the Littoral Combat Ships, LCS-1 and LCS-2. Issues such as motion sickness, motion-induced interruptions with task performance, sleep loss, and fatigue were evaluated during rough water trials. The data were examined for possible inclusion in discrete event model (IMPRINT) of human workload and performance.

 

INVENTORY OPTIMIZATION FOR NAVSUP ENTERPRISE RESOURCE PLANNING 

ResearchersJavier Salmerón, Associate Professor
Emily M. Craparo, Assistant Professor
SponsorNaval Supply Systems Command, Weapons Systems Support
ObjectiveTo develop mathematical optimization models to guide tactical, retail, and wholesale inventory level decisions at Naval Supply Systems Command, Weapons Systems Support (NAVSUP WSS). The ultimate goal is to enhance operational readiness and increase fiscal efficiency by establishing optimal stocking levels for all echelons of supply that NAVSUP WSS manages. We will prioritize our model development in three distinct areas of NAVSUP WSS support: (a) wholesale, (b) readiness-based sparing, and (c) retail, multi-echelon.
SummaryWe began work in November 2013, so progress to date has been modest. The wholesale problem comprises approximately 250,000 items grouped into 250 "Level-Setting Segment Indicator" (LSSI) groups. We have begun the development of an optimization model to minimize (weighted) fill rate deviations with respect to target levels for each LSSI group. The model includes safety stock budgetary constraints, among others. Preliminary testing indicates that we may obtain near-optimal solutions in a reasonable time employing Lagrangian relaxation.

 

OPTIMIZATION OF COMPLEX SYSTEMS 

ResearchersJohannes O. Royset, Associate Professor
SponsorAir Force Office of Scientific Research
ObjectiveWe develop algorithms for solving difficult optimization problems where the objective and/or constraint functions cannot be computed exactly but must be approximated. In particular, we focus on the construction of efficient precision adjustment schemes for controlling the approximations within algorithms.
SummaryThe research is directed towards three classes of optimization problems: (i) stochastic programs where functions are defined in terms of expectations, (ii) semi-infinite programs where functions are nonsmooth max-functions, and (iii) optimal control problems where functions are given by the solution of ordinary and partial differential equations. We have achieved major advances on problems of class (i), i.e., stochastic programs, and obtained results that show the potential for significant computational savings when the precision of approximations is controlled by a discrete-time optimal control problem. We also obtained new rate of convergence results on part (ii) semi-infinite programs and part (iii) control problems.

 

ESTIMATION AND UNCERTAINTY QUANTIFICATION OF UNCERTAIN SYSTEMS 

ResearchersJohannes O. Royset, Associate Professor
SponsorArmy Research Office
ObjectiveWe will carry out a fundamental study of statistical estimation and function approximation and the use of such estimates in uncertainty quantification, rare-event prediction, and information fusion for a broad range of stochastic systems.
SummaryWe propose to develop a flexible framework for estimation of density functions, regression curves, and other quantities that systematically incorporates hard information derived from physics-based sensors, field test data, and computer simulations as well as soft information from human sources and experiences. The project focuses on two main areas: (i) we will consider complex systems subject to random input parameters and will develop epi-spline-based procedures for constructing functional models of the system as well as for estimating probability density functions, moments, quantiles, and rare events of the resulting random system performance; and, (ii) In the context of target detection, tracking, and situational awareness, we will construct epi-spline-based procedures for information fusion of hard data from physics-based sensors with soft contextual information and predictions from human sources pertaining to past, current, and future time periods.

 

ASYMPTOTIC ANALYSIS OF SAMPLE ALLOCATION IN STOCHASTIC OPTIMIZATION 

ResearchersJohannes O. Royset, Associate Professor
Roberto Szechtman, Associate Professor
SponsorAir Force Office of Scientific Research
ObjectiveIn the sample average approximation context, we determine the allocation of the computing budget that leads to the fastest convergence in distribution to an optimal solution.
SummaryMany problems require optimization of a function that can be expressed as the expectation of a function of random variables. In the context of sample average approximation, we study the efficient allocation of the computing budget between two competing demands: generating samples to reduce uncertainty about sample averages, and carrying out iterations of an optimization algorithm. Nonlinear optimization algorithms can typically be characterized by a convergence rate, which affects the resulting allocation. We analyze the case of a single starting point and of multiple starting points, the later being appropriate for optimization of non-convex functions. We find that for linear and super linear convergent optimization algorithms, little effort is directed to algorithm iteration.

 

FUTURE ENERGY NEEDS SIMULATOR

ResearchersPaul J. Sanchez, Lecturer
SponsorMarine Corps Expeditionary Energy Office
Objective
  • Create prototype for the Future Energy Needs Simulator (FERNS) simulation
  • Initiate development of a case study involving a structured investigation of FERNS
SummaryWe developed a prototype model for investigating resource considerations and energy strategy for the Marine Corps. The Future Energy Resource Needs Simulator (FERNS) is based on a framework initially developed to anticipate generic future needs for the armed forces. It is premised on the idea that we cannot know what specific incidents will occur and when, but history tells us there will be a mixture of trigger events such as border incidents, etc., which have had consistent rates of occurrence during the past several decades. Data regarding improvements in energy efficiency and the development of new energy supplies are available. Of interest to E2O are the insights such a model may provide as it matches the needs to the availability of energy and related resources over time. Long run goals for the model include demonstrating how a large-scale simulation experiment can provide better insights about the simulator's key drivers of performance.

 

IMPROVING COST BENEFIT ASSESSMENT FOR ENERGY INITIATIVES USING ROBUST DESIGN

ResearchersSusan M. Sanchez, Professor
SponsorUnited States Marine Corps (USMC) Expeditionary Energy Office (E2O)
ObjectiveTo adapt robust design concepts initially developed for manufacturing models to cost estimation models, with a focus on those associated with expeditionary energy initiatives.
SummaryLife Cycle Cost (LCC) assessments are of interest during the design phase for new systems. These often involve costs that must be estimated from a variety of different sub-models, including cost models constructed from historical data, forecast models that attempt to predict future economic conditions, and economy-of-scale models that impact production schedules, and more. When these disparate models are put together to obtain an overall cost model, many of these individual sources of uncertainty end up being aggregated or ignored. Consequently, the cost estimates may not provide program managers with appropriate assessments of the risk and overall variability of the new systems. We propose a structured approach for obtaining robust LCC estimates by taking into account a broad set of environmental noise conditions. This will enable program managers to better understand the uncertainty in their overall estimates, and to identify any decision factors combinations that result in both low costs and low cost variability. This may provide guidance on which of the many potential uncertainty sources require close monitoring, and which can safely be disregarded. We illustrate this approach with a model the USMC is using for cost/benefit analysis of solar and other alternative energy systems.

 

PERFORMANCE SHAPING FUNCTIONS (PSF) FOR ENVIRONEMTNAL STRESSORS: SLEEP AND PSYCHOMOTOR PERFORMANCE IN MOTION

ResearchersNita Lewis Shattuck, Associate Professor
SponsorOffice of Naval Research Code 34
Objective

The objectives of this work were to participate in the development of procedures for assessing environmental stressors, particularly sleep and psychomotor performance, on human performance in military systems.

The technical objectives were a) to develop a model of how task performance is influenced by the combination of sleep, fatigue, circadian rhythm, and watch-standing schedules; and b) to examine the complex relationship between motion and sleep and the subsequent implications for human performance.

Summary

The LCS trials have been an ongoing effort to look at motion and performance in a shipboard environment. A large team comprised of multiple agencies and universities worked together to assess environmental stressors on two LCS ships. The team members worked together to develop tests and test procedures and to implement those procedures during at-sea trials. The research team communicated through weekly teleconference meetings and periodic in-person sessions to coordinate, develop, and implement a test plan for FY13 and FY14 shipboard testing and analytical strategy.

The approach to the sleep and fatigue issue were to collect data onboard Navy ships using actigraphy to measure sleep and activity levels. The Littoral Combat Ship (LCS 1 Freedom and LCS 2 Independence) were selected as test platforms for the at-sea trials. The calendar year 2013 (CY13) data collection period focused on a calm water environmental setting. Preliminary sleep data on 38 participants was analyzed with results comparing sleep and Switching Task performance. Additional test and study plans were developed for CY14 rough water collection period.

 

HABITABILITY ASSESSMENT TEST

ResearchersNita Lewis Shattuck, Associate Professor
SponsorAmphibious Vehicle Test Branch, USMC
ObjectiveThe objective of this research was to characterize the effect of waterborne movement in amphibious vehicles on the combat effectiveness of embarked infantry. Combat effectiveness is defined by those functions affecting an infantryman's ability to conduct combat operations during an amphibious assault. Although there are many factors related to human performance, this test focused on three functions affected by motion-induced impairment that directly relate to an infantryman's combat effectiveness. These three functions are the ability to move, shoot and communicate.
SummaryWe observed changes in the decision-making skills and reaction times of Marines following two and three hours of waterborne motion exposure. Marine participants in the HAT study also reported consecutively higher levels of motion sickness with increasing length of motion exposure. We observed greater variability in physical coordination completion times following motion exposure. Decisions regarding the length of transit time for amphibious operations should consider these performance changes when analyzing alternatives for amphibious operations.

 

IMPLEMENTATION OF AN ALTERNATIVE WATCHSTANDING SCHEDULE ON USN SURFACE COMBATANTS

ResearchersNita Lewis Shattuck, Associate Professor
SponsorNaval Medical Research Center, Advanced Medical Development Program
ObjectiveEvaluate the effectiveness of an alternative watchstanding schedule for the United States Navy Surface Warfare community.
SummaryFatigue mitigation strategies such as the alternative watchstanding (3/9) study was tested by a field trial conducted upon an operational USN surface combatant ship using USN sailors performing mission relevant tasks. Sailors were outfitted with actiwatches and psychomotor vigilance testing (PVTs) performed to determine sailor's fatigue levels. We implemented a series of studies of crewmembers onboard USN Surface combatants to assess daily activities, sleep patterns, cognitive performance, sailor/command satisfaction surveys regarding alternative watchstanding schedules. Phase I/II of this study was completed onboard a United States destroyer in calendar year 12.

Based on observations and data analysis, additional data was collected onboard the USS Jason Dunham. The most recent data collection period looked at the implementation of the 3/9 alternative watch schedule during operational settings. Data was collected on 122 crewmembers during a three-week underway period while the ship and crew were conducting operations in the Persian Gulf and Arabian Sea. The results indicate that when manning will support a four-section watchbill, the 3/9 watchstanding schedule is a better alternative to the traditional 6/6, 5/15 or 5/10 schedules, providing significantly longer sleep periods with higher sleep quality during the naturally-occurring nighttime period.

The modified 6/18 rotating watch schedule was also looked at onboard the USS Benfold as a possible alternative to the traditional schedules. The results showed reduced sleep quantities and degraded performance among watchstanders.

Phase III of the 3/9 watch schedule studies continued onboard the USS Nimitz during calendar year 2013 (CY13). This data collection focused on the sleep patterns of Sailors in the Reactor Department on the carrier working a traditional watchbill. Plans to collect data on the Reactor Department following a circadian watchbill are in progress for CY14.

 

WORK-REST PATTERNS IN OPERATIONAL ENVIRONMENTS: PREVENTING STRESS-RELATED INJURIES IN THE U.S. NAVY

ResearchersNita Lewis Shattuck, Associate Professor
SponsorOPNAV N171 21st Century Sailor Office
ObjectiveDetermine if there are differences in mood, reported stress levels, alertness/vigilance levels and sleep patterns of Sailors using an alternative 3/9 hour watchstanding schedule as compared to conventional watchstanding schedules. This assessment was conducted during a deployment period onboard Navy surface combatants during operational conditions.
SummaryFatigue mitigation strategies such as the alternative watchstanding (3/9) study was tested by a field trial conducted upon an operational USN surface combatant ship using USN sailors performing mission relevant tasks. Sailors were outfitted with actiwatches and psychomotor vigilance testing (PVTs) performed to determine sailor's fatigue levels. We implemented a series of studies of crewmembers onboard USN Surface combatants to assess daily activities, sleep patterns, cognitive performance, sailor/command satisfaction surveys regarding alternative watchstanding schedules. In this ongoing study, we will compare multiple ships and departments in comparable operational conditions. Phase I/II/III of this study was completed onboard two United States destroyers and a Carrier in calendar year 2012 (CY12) and CY13. Other test platform ships were identified during CY12. The laboratory phase of the 3/9 watch schedule studies was developed and planned for CY14 execution. This phase will look to validate the 3/9 watch schedule in a controlled laboratory setting.

 

THE SCIENCE OF TEST 

ResearchersRachel T. Silvestrini, Assistant Professor
SponsorAir Force Institute of Technology
ObjectiveExpand the use of and the research in experimental design techniques to Department of Defense (DoD) Test and Evaluation Programs.
SummaryThe Air Force Institute of Technology (AFIT) and the Naval Postgraduate School (NPS) have partnered with Arizona State University and Virginia Tech to form a consortium to support graduate research in the area of test science, specifically related to areas of experimental design and the statistical analysis of those designs and the area of live, virtual, constructive test integration into test and evaluation. This team has been specifically designed and developed to address the research challenges facing the DoD Test Enterprise in leveraging the scientific capabilities that exist and can be developed to support the defense test community.

 

SEXUAL ASSAULT DATA ANALYTICS 

ResearchersLyn R. Whitaker, Associate Professor
SponsorHQMC Sexual Assault Prevention & Response, Personal & Family Readiness Division
ObjectiveThis work is part of the NPS-OPNAV Studies Program. The Commandant of the Marine Corps has led the effort Corps wide on Sexual Assault Prevention and Response (SAPR). The Marine Corps SAPR office (SAPEO) has implemented many new training and briefing tools in support of this effort. The number of reported assaults has increased in the past few years. This increase may represent a greater propensity of Marine Corp victims to report sexual assault. To understand the increase and to determine progress, a baseline needs to be established. Augmenting the newly constructed Defense Sexual Assault Incident Database (DSAID) used by SAPRO to compile their annual reports, with historical Marine Corp wide personnel data from Defense Manpower Data Center (DMDC) and pertinent survey data, our objective is to establish a process for determining a baseline, provide a deeper understanding of the factors related to increased reporting, and insight into what information needs to be collected to improve analysis based of extant data.
SummaryThe funding for work to be done in the fall of calendar year 2013 quarter arrived too late in that quarter to use it. It also arrived too late to fund travel for Capt. Bethany Kaufman to spend part of her experience tour at the USMC SAPR Office. We were able to get Institutional Review Board IRB approval to procure DSAID data from SAPRO and Capt. Kaufman was able to make progress in obtaining that data.

 

DEFENDING INTERDEPENDENT INFRASTRUCTURE SYSTEMS

ResearchersR. Kevin Wood, Distinguished Professor
Javier Salmerón, Associate Professor
David L. Alderson, Associate Professor
SponsorDefense Threat Reduction Agency through the University of Texas at Austin
ObjectiveThis research, in collaboration with Professor Ross Baldick of the University of Texas at Austin, seeks to develop new theory, models and algorithms for optimal design or retrofit of interdependent infrastructure systems. The goal is to make such systems resilient to kinetic and other types of WMD attacks.
SummaryWe have developed new techniques for assessing the resilience to attack of interdependent infrastructure systems through attacker-defender (AD) models, and for optimizing that resilience through defender-attacker-defender (DAD) models. A general algorithm for solving DAD models was tested successfully on full-scale model of the traffic flow in the highway network in the San Francisco Bay Area. We have also finished theoretical and computational development of an AD model for analyzing a strategic spare-parts policy for an electric power grid. The second-stage of this model optimally allocates spare parts after an attack, and then operates the power grid optimally, over time, as standard repairs take place. Full-scale models solve in a few hours. We have also developed a new enumerative method for solving the mixed-integer problems that arise when solving AD and DAD problems using the preferred technique of "global Benders decomposition." Initial testing indicates that order-of-magnitude computational speedups are possible in some cases, with parallel implementations pushing the advantage even farther.

This is an official U.S. Navy website.
All information contained herein has been approved for release by the NPS Public Affairs Officer.
Page Last Updated: Apr 1, 2014 10:02:07 AM | Contact the Webmaster