Search

Purpose: Lack of diagnostic and prognostic biomarkers in hepatocellular carcinoma impedes stratifying patients based on their risk of developing cancer. The aim of this study was to evaluate phenotypic and genetic heterogeneity of circulating epithelial cells (CECs) based on asialoglycoprotein receptor 1 (ASGR1) and miR-122-5p expression as potential diagnostic and prognostic tools in patients with hepatocellular carcinoma (HCC) and liver cirrhosis (LC). Methods: Peripheral blood samples were extracted from LC and HCC patients at different disease stages. CECs were isolated using positive immunomagnetic selection. Genetic and phenotypic characterization was validated by double immunocytochemistry for cytokeratin (CK) and ASGR1 or by in situ hybridization with miR-122-5p and CECs were visualized by confocal microscopy. Results: The presence of CECs increased HCC risk by 2.58-fold, however, this was only significant for patients with previous LC (p = 0.028) and not for those without prior LC (p = 0.23). Furthermore, the number of CECs lacking ASGR1 expression correlated significantly with HCC incidence and absence of miR-122-5p expression (p = 0.014; r = 0.23). Finally, overall survival was significantly greater for patients at earlier cancer stages (p = 0.018), but this difference was only maintained in the group with the presence of CECs (p = 0.021) whereas progression-free survival was influenced by the absence of ASGR1 expression. Conclusion: Identification and characterization of CECs by ASGR1 and/or miR- 122-5p expression may be used as a risk-stratification tool in LC patients, as it was shown to be an independent prognostic and risk-stratification marker in LC and early disease stage HCC patients.

Purpose: Lack of diagnostic and prognostic biomarkers in hepatocellular carcinoma impedes stratifying patients based on their risk of developing cancer. The aim of this study was to evaluate phenotypic and genetic heterogeneity of circulating epithelial cells (CECs) based on asialoglycoprotein receptor 1 (ASGR1) and miR-122-5p expression as potential diagnostic and prognostic tools in patients with hepatocellular carcinoma (HCC) and liver cirrhosis (LC). Methods: Peripheral blood samples were extracted from LC and HCC patients at different disease stages. CECs were isolated using positive immunomagnetic selection. Genetic and phenotypic characterization was validated by double immunocytochemistry for cytokeratin (CK) and ASGR1 or by in situ hybridization with miR-122-5p and CECs were visualized by confocal microscopy. Results: The presence of CECs increased HCC risk by 2.58-fold, however, this was only significant for patients with previous LC (p = 0.028) and not for those without prior LC (p = 0.23). Furthermore, the number of CECs lacking ASGR1 expression correlated significantly with HCC incidence and absence of miR-122-5p expression (p = 0.014; r = 0.23). Finally, overall survival was significantly greater for patients at earlier cancer stages (p = 0.018), but this difference was only maintained in the group with the presence of CECs (p = 0.021) whereas progression-free survival was influenced by the absence of ASGR1 expression. Conclusion: Identification and characterization of CECs by ASGR1 and/or miR- 122-5p expression may be used as a risk-stratification tool in LC patients, as it was shown to be an independent prognostic and risk-stratification marker in LC and early disease stage HCC patients.

Purpose: Lack of diagnostic and prognostic biomarkers in hepatocellular carcinoma impedes stratifying patients based on their risk of developing cancer. The aim of this study was to evaluate phenotypic and genetic heterogeneity of circulating epithelial cells (CECs) based on asialoglycoprotein receptor 1 (ASGR1) and miR-122-5p expression as potential diagnostic and prognostic tools in patients with hepatocellular carcinoma (HCC) and liver cirrhosis (LC). Methods: Peripheral blood samples were extracted from LC and HCC patients at different disease stages. CECs were isolated using positive immunomagnetic selection. Genetic and phenotypic characterization was validated by double immunocytochemistry for cytokeratin (CK) and ASGR1 or by in situ hybridization with miR-122-5p and CECs were visualized by confocal microscopy. Results: The presence of CECs increased HCC risk by 2.58-fold, however, this was only significant for patients with previous LC (p = 0.028) and not for those without prior LC (p = 0.23). Furthermore, the number of CECs lacking ASGR1 expression correlated significantly with HCC incidence and absence of miR-122-5p expression (p = 0.014; r = 0.23). Finally, overall survival was significantly greater for patients at earlier cancer stages (p = 0.018), but this difference was only maintained in the group with the presence of CECs (p = 0.021) whereas progression-free survival was influenced by the absence of ASGR1 expression. Conclusion: Identification and characterization of CECs by ASGR1 and/or miR- 122-5p expression may be used as a risk-stratification tool in LC patients, as it was shown to be an independent prognostic and risk-stratification marker in LC and early disease stage HCC patients.

Purpose: Lack of diagnostic and prognostic biomarkers in hepatocellular carcinoma impedes stratifying patients based on their risk of developing cancer. The aim of this study was to evaluate phenotypic and genetic heterogeneity of circulating epithelial cells (CECs) based on asialoglycoprotein receptor 1 (ASGR1) and miR-122-5p expression as potential diagnostic and prognostic tools in patients with hepatocellular carcinoma (HCC) and liver cirrhosis (LC). Methods: Peripheral blood samples were extracted from LC and HCC patients at different disease stages. CECs were isolated using positive immunomagnetic selection. Genetic and phenotypic characterization was validated by double immunocytochemistry for cytokeratin (CK) and ASGR1 or by in situ hybridization with miR-122-5p and CECs were visualized by confocal microscopy. Results: The presence of CECs increased HCC risk by 2.58-fold, however, this was only significant for patients with previous LC (p = 0.028) and not for those without prior LC (p = 0.23). Furthermore, the number of CECs lacking ASGR1 expression correlated significantly with HCC incidence and absence of miR-122-5p expression (p = 0.014; r = 0.23). Finally, overall survival was significantly greater for patients at earlier cancer stages (p = 0.018), but this difference was only maintained in the group with the presence of CECs (p = 0.021) whereas progression-free survival was influenced by the absence of ASGR1 expression. Conclusion: Identification and characterization of CECs by ASGR1 and/or miR- 122-5p expression may be used as a risk-stratification tool in LC patients, as it was shown to be an independent prognostic and risk-stratification marker in LC and early disease stage HCC patients.

Naval Supply Systems Command Weapon Systems Support currently uses Distributed Requirement Planning (DRP) to generate repair and procurement quantity recommendations for Supply Class IX repair parts. In an effort to provide an optimization-based method for making wholesale inventory decisions, a mixed integer linear optimization program termed Wholesale Inventory Optimization Model (WIOM) was developed as an alternative to existing tools. In its current form, however, WIOM does not explicitly consider the repair cycle for those national item identification numbers that may be either procured or repaired. The thesis develops a next-event, discrete-event inventory simulation model, which we term Comparative Replenishment Stream Simulation (COMPRESS) to compare the effects of having two supply streams with differing lead times versus what we commonly see in a single supplier stream. With WIOM strictly considering a “composite” supply stream, the determination needs to be made if the disparity between inventory systems that utilize either a composite or repair-and-procure supply stream is great enough to warrant a WIOM reformulation. We select fill rate as the performance metric to compare these inventory systems. Through COMPRESS we find that there is considerable difference in the resulting fill rates. Additionally, we find that as the variability in demand and lead time increases, the difference in the resulting fill rate values increases as well., http://archive.org/details/comparingnavalin1094562749, Lieutenant Commander, United States Navy, Approved for public release; distribution is unlimited.

During December 2017, the Los Angeles County Fire Department (LACoFD) responded to an unprecedented number of wildland fires spanning thousands of square miles within the county. While most of the fires were successfully contained, several caused widespread catastrophic damage. During red flag (high-risk) days, LACoFD currently uses augmented staffing, either by moving on-duty equipment and personnel, or mobilizing those who are off duty to reduce response time. Operational duty chiefs make these augmentation decisions based on current weather conditions and experience. This thesis develops regression models to estimate the probability and potential burned acreage of wildland fires in each of 21 sub-areas in the county. Then, a budget-constrained optimization model reassigns resources between sub-areas in order to minimize expected population displacement due to wildland fire. A comparison of these automated techniques with those manual decisions made during December 2017 reveals significant improvements to augmented staffing that can be made at a lower cost.

Naval Supply Systems Command Weapon Systems Support currently uses Distributed Requirement Planning (DRP) to generate repair and procurement quantity recommendations for Supply Class IX repair parts. In an effort to provide an optimization-based method for making wholesale inventory decisions, a mixed integer linear optimization program termed Wholesale Inventory Optimization Model (WIOM) was developed as an alternative to existing tools. In its current form, however, WIOM does not explicitly consider the repair cycle for those national item identification numbers that may be either procured or repaired. The thesis develops a next-event, discrete-event inventory simulation model, which we term Comparative Replenishment Stream Simulation (COMPRESS) to compare the effects of having two supply streams with differing lead times versus what we commonly see in a single supplier stream. With WIOM strictly considering a “composite” supply stream, the determination needs to be made if the disparity between inventory systems that utilize either a composite or repair-and-procure supply stream is great enough to warrant a WIOM reformulation. We select fill rate as the performance metric to compare these inventory systems. Through COMPRESS we find that there is considerable difference in the resulting fill rates. Additionally, we find that as the variability in demand and lead time increases, the difference in the resulting fill rate values increases as well.

During December 2017, the Los Angeles County Fire Department (LACoFD) responded to an unprecedented number of wildland fires spanning thousands of square miles within the county. While most of the fires were successfully contained, several caused widespread catastrophic damage. During red flag (high-risk) days, LACoFD currently uses augmented staffing, either by moving on-duty equipment and personnel, or mobilizing those who are off duty to reduce response time. Operational duty chiefs make these augmentation decisions based on current weather conditions and experience. This thesis develops regression models to estimate the probability and potential burned acreage of wildland fires in each of 21 sub-areas in the county. Then, a budget-constrained optimization model reassigns resources between sub-areas in order to minimize expected population displacement due to wildland fire. A comparison of these automated techniques with those manual decisions made during December 2017 reveals significant improvements to augmented staffing that can be made at a lower cost., Office of Naval Research, http://archive.org/details/optimizingresour1094562835, Outstanding Thesis, Ensign, United States Navy, Approved for public release; distribution is unlimited.

Aviation Depot Maintenance Throughput Optimization

Project Summary: In 2004, the Naval Aviation Enterprise transitioned to a system of Performance/Pricing Models (P/PM) in an effort to support decision makers and gain efficiency in its funding activities. P/PMs consist of two key elements: driver-based performance management, which identifies the organization’s desired outcomes and the drivers that have quantifiable impacts on them, and predictive analytics, which identify high-leverage drivers and forecast performance. The F/A-18 Super Hornet has been the workhorse of the Navy and Marine Corps' tactical aviation squadrons for more than 30 years. However, over the last decade, the amount of hours flown in a mission capable (MC) status has steadily declined despite increases in funding and the utilization of the P/PMs to better allocate that funding. This work, in conjunction with efforts by the Chief of Naval Operations Assessment Division (N81) to assess the causes of this readiness decay, consists of a mixed methodology analysis of the following F/A-18 P/PMs: Flight Hour Program (FHP), Spares Model, and the Depot Readiness Assessment Model (DRAM) suite—including the Engine Depot Readiness Assessment Model (EDRAM) and the Airframe Depot Readiness Assessment Model (ADRAM). The seminal product of this research is a logical framework describing the relationships between the P/PMs with a particular focus on the model drivers determined as key to model performance. The framework serves as a foundation for future work in understanding the complexities of the Navy's aviation problem set. This research also illuminates the extent to which human decisions shape model outputs.

Naval Supply Systems Command Weapon Systems Support (NAVSUP WSS) serves as the Navy’s inventory control point, managing approximately 375,000 line items. Constrained by funding, NAVSUP WSS uses the Wholesale Inventory Optimization Model (WIOM), a mixed-integer linear program developed by Naval Postgraduate School faculty, to maximize customer service. Since demand distributions for different parts change over time, NAVSUP WSS updates the inputs to WIOM and reruns it quarterly. However, large changes to the solution create an administrative burden. To deal with this problem, referred to as churn, WIOM has a persistence parameter that can discourage change from one run to the next, but it is inherently at odds with customer service performance. This thesis presents a new model, the Comparative Optimized Results Simulation (CORS). Using CORS, the thesis explores the system’s performance under different settings of the persistence parameter and different periodicities of running WIOM. The thesis finds that periodicities greater than quarterly significantly degrade customer service. Additionally, the thesis finds that increasing the persistence parameter dramatically improves churn while only marginally degrading customer service.

The Personal Gear Issue (PGI) Division at each Expeditionary Support Unit is responsible for ordering, inventory management, and distribution of all personal gear issued. This work adapts an existing optimization model to determine appropriate ordering policies, specifically, the order point, s, and order quantity, Q, for each line item. The model seeks to minimize cost of inventory subject to target fill rate for each item, maximum number of orders per month, and other constraints. We consider realistic conditions such as current demand, 90 orders per month, and 99% fill rate. Under these conditions, the optimal s and Q values result in a significant reduction in the cost of inventory, from $8.2M (current PGI’s inventory value) down to $1.23M. In addition, our model shows, via sensitivity analysis, the effect that varying fill-rate targets, monthly order limit, and demand have on cost. For example, we observe a nonlinear increase in cost as a function of demand increase. We recommend that PGI Division revises its ordering policy to include formal optimization analysis and to periodically track demand in order to re-evaluate the optimal s and Q values., http://archive.org/details/expeditionarylog1094560392, Lieutenant Commander, United States Navy, Approved for public release; distribution is unlimited.

The Marine Corps manages a centralized secondary reparable inventory of over 1,000 different items with an enterprise-wide retail value of more than $600 million. These secondary reparable items are those components and subcomponents deemed economical to repair based on their expected costs and life cycles. The current ad hoc procedure to determine how to replenish each damaged secondary reparable relies on coordination between multiple organizations using information from various sources and varies from location to location. Ongoing Marine Corps efforts aim to determine the effectiveness of the Centralized Secondary Reparable Management Program by analyzing the cost and quality of these repairs and assessing the current manning of Marine maintenance activities. This thesis develops optimization models to assist in the determination of how to conduct these component repairs in the timeliest and least costly manner. Results from the model demonstrate the value of the organizations currently integrated into the secondary repairable maintenance cycle. Additionally, the solution provides strong evidence to support repair policies based on national stock numbers and location, drastically reducing the complexity of the current ad hoc procedures.

The Marine Corps manages a centralized secondary reparable inventory of over 1,000 different items with an enterprise-wide retail value of more than $600 million. These secondary reparable items are those components and subcomponents deemed economical to repair based on their expected costs and life cycles. The current ad hoc procedure to determine how to replenish each damaged secondary reparable relies on coordination between multiple organizations using information from various sources and varies from location to location. Ongoing Marine Corps efforts aim to determine the effectiveness of the Centralized Secondary Reparable Management Program by analyzing the cost and quality of these repairs and assessing the current manning of Marine maintenance activities. This thesis develops optimization models to assist in the determination of how to conduct these component repairs in the timeliest and least costly manner. Results from the model demonstrate the value of the organizations currently integrated into the secondary repairable maintenance cycle. Additionally, the solution provides strong evidence to support repair policies based on national stock numbers and location, drastically reducing the complexity of the current ad hoc procedures., LX, I&L, HQMC, http://archive.org/details/optimizingmarine1094559573, Outstanding Thesis, Captain, United States Marine Corps, Approved for public release; distribution is unlimited.

We develop techniques to optimise the locations and surveillance scheduling of tower-mounted camera systems used by a military force in an urban setting. Using a game-theoretic foundation, we seek to minimise expected damage from attacks or other adversarial events (e.g., emplacements of improvised explosive devices). Assuming that at most one camera may surveil a single point of interest (POI) at any time, a mixed-integer program uses an additive-probability model to optimise the placement of towers, while allocating ‘aggregate, normalised surveillance time’ between cameras and POIs. Linear-programming-based column generation then creates a probability distribution for camera-to-POI assignments to define implementable schedules. We prove that such schedules must exist, making the additive probability model exact. Computational examples on realistically sized problems produce high-quality solutions quickly, with quality suffering only when the number of cameras available nears the number of POIs to be surveilled. We show that an alternative game-theoretic model may produce better solutions when such a situation arises.

Naval Supply Systems Command (NAVSUP) supports Navy, Marine Corps, Joint and Allied Forces with their inventory of more than 430,000 items worth $21 billion using several distribution sites. Choosing the optimum order-point and order-quantity for each item is important to meet the stochastic demand while satisfying multiple restrictions such as budget and maximum number of orders. The Site Demand-Based Level Inventory Optimization Model (SIOM) is a mixed-integer, linear program developed at the Naval Postgraduate School to provide NAVSUP planners with guidance on this complex problem. Ongoing tests have been successful, but SIOM's computational run times are long. This thesis introduces a new, faster reformulation (SIOMsQ) that approximates the solution of the same problem by reducing the possible candidate sets of order-points and order-quantities for each item. We find that the solutions suggested by SIOMsQ are better than or very close to those of SIOM in test cases provided by NAVSUP, with substantially shorter computational times. Therefore, we recommend using SIOMsQ versus SIOM.

The Navy Supply Systems Command holds $21 billion of inventory across 430,000 repair part line items in support of ongoing naval operations. Selecting the correct reorder point for each of those line items requires a careful balance between tying up precious funding by holding too much inventory that may not be issued, and not holding enough inventory to meet customer demands in a timely fashion. This thesis compares three different methods for selecting the reorder point. The first method is a calculation that offers no attempt at optimization but is simple to understand. The second method is provided by a contractor and uses some optimization with unknown algorithmic details. The last method is a mixed-integer, linear optimization model. Comparative Inventory Simulation, a discrete event simulation model, is designed to find fill rates achieved for each National Item Identification Number under the different reorder methods. We find that average fill rates are higher in 22 out of 24 cases, and average backorder lengths are up to 50% shorter, using the last method.

Marine Corps Installation and Logistics Command is seeking assistance to improve operations within Marine Corps Maintenance Production plants. The problem addressed in this thesis deals with production lines: there must be a proper balance of parts on hand and inventory costs to ensure optimal production output. This problem becomes increasingly difficult to solve as production-line complexity increases and overall budget flexibility decreases. As the Marine Corps enters a time of fiscal austerity and reduced overseas combat operations, it is critical to optimize its processes so major end items are refurbished in the quickest and most cost-effective manner, thereby ensuring maximum combat effectiveness. This research focuses on developing a proof of concept analytical tool to better facilitate order management of repair parts. This tool integrates optimization and discrete-event simulation. This construct can potentially provide an effective means in improving order management decisions. However, the effectiveness of the tool is contingent on accurate vehicle condition history, parts order history, and/or future estimated parts shipping dates. Information derived from the analysis can be used to make recommendations for reorder policy, enable future model development, and improve the overall maintenance production process.

To support the U.S. Administration’s announced rebalancing, or pivot, to the Asia-Pacific region, the U.S. Marine Corps is pushing its footprint beyond established logistics support locations. Yet, the Corps’ ability to rapidly respond to primary missions, such as international military cooperation, humanitarian assistance, and defense of U.S. interests, must be maintained. This Marine diaspora from Okinawa, its current concentration, must consider the logistical vagaries of time, space, and expense. This thesis develops an optimization-based tool by which item-specific details are combined with theater logistical constraints to analyze present logistical performance and what-if capability for future troop redeployments within the region. The mathematical model is applied to optimally allocate Class IX repair parts for the Marine Corps across the Pacific theater. In particular, the combined relationship of Defense Logistics Agency and the Marine’s intermediate level supply units is analyzed. The optimization model minimizes a logistical system cost of multiple priority demands from disparate locations. The model is flexible enough to allow any number of stock items, additional user locations, and supply nodes at operational and strategic levels. While the focus is the force rebalancing in the Pacific, the intent is a general tool that can support other theaters.

The Littoral Combat Ship (LCS) is a naval combatant designed to operate in the littoral regions. Twenty-four LCSs will be built over the next five years employing a crew rotation concept where three crews rotate between two ships. During the construction period, an experienced crew must be assigned, which disrupts the desired crew rotation in ships already built. This thesis develops LCS Scheduler (LCSS), a mathematical optimization model using a mixed-integer, linear program (MIP) to aid in assigning LCS crews to LCS ships. LCSS’s objective is to minimize the penalty associated with assigning crews outside of their desired ship pairing and/or extending them beyond four months in a phase. Results are compared based on solve time and penalty value. The MIP solution has the best quality. Yet, even for a shorter-than-desired time horizon, it takes many hours of computation. Rolling horizon is a heuristic approach that produces a full, long-term schedule in under an hour but requires manual modifications to misaligned crews. Fix-and-relax is a more-elaborate heuristic with potential benefits to crew alignment for longer-range schedules. The planner must balance solve time and solution quality when determining the approach to LCSS.

Taiwan is prone to many natural disasters, especially typhoons. This thesis adapts an existing stochastic prepositioning optimization model to create a tool for Taiwan military disaster recovery planners, and then uses experimental design techniques to systematically explore solutions. The goals are to minimize the expected number of casualties and unmet commodities demands, and to determine the average number of workers deployed in response to each scenario. A design of experiments methodology is applied to the optimization model to reveal how uncertainty in the parameters translates to uncertainty in objective function values. The approach can also identify the parameters with the greatest impact on the objective function, and result in more robust solutions. The analysis demonstrates that it is not always necessary to spend as much money and deploy as many workers as in the past in order to get the best results. Additionally, the approach shows how a decision maker, with more accurate and current weather reports, can refer to the path and intensity of typhoons while making rescue plans. In summary, this research shows that there is great potential for quantitative methods to improve the disaster-relief planning process.

To support the U.S. Administration’s announced rebalancing, or pivot, to the Asia-Pacific region, the U.S. Marine Corps is pushing its footprint beyond established logistics support locations. Yet, the Corps’ ability to rapidly respond to primary missions, such as international military cooperation, humanitarian assistance, and defense of U.S. interests, must be maintained. This Marine diaspora from Okinawa, its current concentration, must consider the logistical vagaries of time, space, and expense. This thesis develops an optimization-based tool by which item-specific details are combined with theater logistical constraints to analyze present logistical performance and what-if capability for future troop redeployments within the region. The mathematical model is applied to optimally allocate Class IX repair parts for the Marine Corps across the Pacific theater. In particular, the combined relationship of Defense Logistics Agency and the Marine’s intermediate level supply units is analyzed. The optimization model minimizes a logistical system cost of multiple priority demands from disparate locations. The model is flexible enough to allow any number of stock items, additional user locations, and supply nodes at operational and strategic levels. While the focus is the force rebalancing in the Pacific, the intent is a general tool that can support other theaters.

The U.S. Marine Forces Reserve (USMCR, MARFORRES) is conducting realignment studies where discretionary changes may benefit from formal mathematical analysis. This study has developed an optimization tool to guide and/or support Commander, MARFORRES (CMFR) decisions. A prototype of the optimization tool has been tested with data from the units and Reserve Training Centers (RTCs) in the San Francisco, CA and Sacramento, CA areas.

This research creates Marine Corps Reserve Optimizer (MCRO), an optimization tool to aid Marine Forces Reserve (MARFORRES) in the task of geographically situating their subordinate units with respect to demographics. It implements an integer linear program that selects optimal locations for all candidate moving units based on the projected availability of qualified recruits in candidate areas. MCRO optimizes to (a) minimize a penalty function that measures stress with respect to demographics, and (b) minimize unit movement. Two base cases are demonstrated, one illustrating the total demographic stress with 2011 population data without allowing unit movements, and another with the projected 2036 population under the same conditions. We then allow MCRO to recommend movements, and find that (i) the relocation of 10 units reduces the number of areas experiencing the highest penalty from nine to three, and (ii) all stress can be relieved in 56 movements. Finally, we use MCRO to evaluate and quantify the demographic impact of four possible unit movements MARFORRES is currently considering.

Weapon systems are comprised of parts that are subject to random failures. When a part fails, it must be replaced by an operable part that is provided by a supply network that supports the system. Supply networks consist of many locations where spare parts are held, known as echelons. Examples include depots, fulfillment centers, and customers. When many identical weapon systems operate in parallel and rely on a multi-echelon supply network for replacement parts, decision makers must choose where and how to invest their resources into the purchase of spare parts. This thesis uses stochastic optimization to leverage those decisions in order to maximize the expected number of time periods a set of weapon systems are available for use. Specifically, we develop a model that determines the optimal stock levels of spare parts to store at each echelon of the supply network. The formulation integrates part failure uncertainty, transit times, and monetary constraints. Model outputs also provide decision makers with a clear estimate of marginal availability gains for each dollar invested in purchasing spare parts.

Every few months each United States submarine must return to a port to undergo major maintenance that cannot be conducted at sea. These maintenance periods are called Continuous Maintenance Availability (CMAV) periods. All CMAV scheduling aboard the two remaining submarine tenders in the United States fleet, the USS Emory S. Land (AS 39) and the USS Frank Cable (AS 40), is currently done manually. The schedulers rely on their experience and sound judgment with the goal of successfully completing the most maintenance as quickly and efficiently as possible for approximately 200 jobs, 50 maintenance shops and a host of other considerations. In this thesis, we develop a job-shop scheduling model, the CMAV Scheduler (CMAV-S). This is a large-scale, mixed-integer, linear program that accounts for a variety of scheduling inputs commonly used by planners: job priority, duration, allowed window of execution, prerequisites, mandatory character, workforce used and available (by shop), and special submarine conditions (active or inactive) needed to perform a job. CMAV-S produces near-optimal schedules that achieve maximum value for all scheduled jobs in about one minute. When compared to our own manual scheduling, we observe CMAV-S improves up to 25% the required CMAV length to schedule all maintenance.

The daily flight schedule at Training Air Wing-Two (TW-2) is built manually each day by squadron scheduling officers (SKEDSOs). They rely on their intuition, experience and sound judgment to output a flight schedule. Each SKEDSO spends eight hours a day on this task, but currently there is no measure of the efficiency a given flight schedule. Our goal is to enhance the current planning process by alleviating many of the tedious tasks through an automated optimization program. To that end, this research develops Flight Training Scheduler (FTS), an optimization-based tool, to aid the SKEDSO in production of daily flight schedules. FTS allows the SKEDSO to place an objective, value-oriented metric on the total events scheduled. A typical instance of this problem for TW-2's Phase II students consists of approximately 30 students, 65 flight events and 35 instructor pilots. FTS provides fast, automated guidance to the SKEDSOs that can help them increase throughput of students in the advanced strike training syllabus.

This paper incorporates, as part of an attacker-defender (AD) model for an electric power transmission grid, an inventory of "recovery spares" for high-voltage transformers (HVTs). In this sequential-game model, an attacker first uses limited resources to disable grid components, seeking to maximize second-stage "total cost." This value corresponds to a defender who operates the damaged grid to minimize generating and economic load-shedding costs. The defender's problem includes 1) optimal power-flow solutions to represent post-attack operations, 2) optimized replacement of disabled HTVs with quickly installable recovery spares, and 3) longer-term replacements using new procurements. Global Benders decomposition, with a mixed-integer subproblem, solves the AD model; new enumerative techniques help solve the decomposition's master problem and subproblems. Computational tests demonstrate tradeoffs between the number of recovery spares and worst-case total cost, and show how the model could help guide an inventory strategy for recovery spares in an adversarial setting.

In this thesis, we develop a mixed-integer, linear optimization model to guide the resourcing of submarine maintenance conducted by the U.S. Navys two submarine tenders in the Fifth and Seventh Fleets. We assume maintenance demands are known over a given planning horizon, e.g., one month. Inputs to the model include travel times and costs for fly-away teams and tenders to move to where the maintenance is needed. Each maintenance demand can be divided into tasks with characteristics such as: whether or not tender presence is required; the estimated total number of worker-days required; the maximum number of workers that can simultaneously work on each task; the types of maintenance workers that can perform the task; and task due dates. The models output determines the assignment of personnel to meet the demand at minimum cost, including delay penalties. It also guides personnel travel (as a fly-away team or by tender). In addition, the model can be used to accommodate emergent, unscheduled demands by producing an updated schedule that minimizes the impact on an existing schedule. We test our model on small and realistically sized notional examples to demonstrate the input and output of the models, as well as computational run-times.

This thesis formulates and solves a mixed-integer program to plan the recovery of an electrical power transmission grid that has been damaged by a natural disaster or terrorist attack. The damage can be extensive and recovery can take weeks or months. An efficient recovery plan that maximizes the utilization of repair resources can help ensure swift restoration of services. The network recovery-planning model is implemented in GAMS (General Algebraic Modeling System) and uses CPLEX as the solver. An electrical grid based on IEEE's 300-bus transmission network is used for testing. To simulate varying degrees of damage to the network, we choose up to 20% of the grid's lines, buses and transformers to be placed out of service. Based on the availability of repair resources, the repair-time horizon and penalties for unserved demand, the model produces a repair schedule that minimizes the cost of power shed. We demonstrate that for a network with up to 8% of its components damaged, the model can produce an optimal recovery plan within 20 minutes on a 2 GHz personal computer. For our largest test-case with 20% of network components damaged, the recovery plan is within 7% of optimal after 1 hour of solver time., http://archive.org/details/optimizedrecover109452859, Outstanding Thesis, Approved for public release; distribution is unlimited.

This thesis describes three specialized branch-and-bound (B and B) algorithms for solving a mixed-integer program (MIP) that incorporates standard big-M constructs. The goal is to identify valid values for M that also lead to short solution times. One algorithm initializes large instances of M (giving a weak relaxation of the MIP), and decreases these as required to increase efficiency of the standard B and B. Two algorithms initialize small and possibly invalid instances of M, and subsequently increase those values in an attempt to ensure solution validity. Each algorithm requires a model-specific test condition to detect weak or invalid Ms. We test all algorithms on an uncapacitated k-median problem (a variant of the uncapacitated facility location problem), and one algorithm on a shortest-path interdiction problem (SPIP). We observe substantial reduction in run times in almost all cases tested. When solving for exact solutions, computational results show that the proposed algorithms may reduce solution times by up to 75 per cent for the uncapacitated k-median problem and 99 per cent for the SPIP. When the algorithms yield marginally suboptimal solutions, substantial solution-time improvements are also recorded. While testing is limited, this thesis serves as a proof-of-concept that the proposed adaptive algorithms can be effective in reducing solution times and producing optimal or nearly optimal solutions.

three categories of money. CIPA APU tracks average aircraft age, expected attrition, and allows planners to specify a platformαs mission suitability. It also provides expeditious feedback to requests for alternate scenario feasibility and budget impact. We demonstrate CIPA APU capabilities using a few realistic scenarios., http://archive.org/details/optimizedprocure109456009, US Navy (USN) author

CHDS State/Local, In large-scale domestic disasters, a temporal gap frequently develops between the exhaustion of state and local resources and the arrival of federal resources. To date, strategies for reducing this so-called βgap of painγ have not been based upon scientific methodology. This thesis reviews four alternatives for ensuring continuous availability of critical commodities: pre-positioning, pre-emptive federal action, time-phased deployment, and surge transportation. For a given scenario, the optimum approach is likely to be some combination of these alternatives. Stochastic modeling using optimization techniques holds great promise for producing efficient and effective strategic solutions. This thesis evaluates one such model using two notional scenarios affecting the Washington, D.C. metropolitan area: a Category 4 hurricane and a one-kiloton nuclear explosion near the city center. The results reinforce the validity of using this method to generate viable strategic alternatives for consideration by senior decision-makers. With additional development and testing, the model may be productively applied to a range of natural and man-made incidents, in disparate locations., http://archive.org/details/reducinggapofpai109453597, Chief of Operations Law Division - Major (Maj), Operations Law Division - Joint Force Headquarters-National Capital Region (JFHQ-NCR) - Department of Defense - US Northcom, US Air Force (USAF) author.

Current military leadership is directing the U.S. Navy to engage in theater security cooperation activities or missions to bolster confidence and build trust relationships with other national military forces. Using ships efficiently by maximizing the value of missions accomplished in any time period is important because these missions affect world security, as well as our own national security. Recent research has developed a mixed-integer programming optimization model called Central-West Africa Resource and Mission Allocation (CARMA) that seeks to efficiently route a single naval vessel embarked with expeditionary partnership teams conducting theater security cooperation missions, to maximize the total mission value. The two current algorithms for solving CARMA require commercial software to solve the associated optimization models. This thesis develops a custom-built, license-free heuristic algorithm that provides decent solutions to CARMA in a fraction of the time of these algorithms. The developed heuristic uses limited enumeration to generate feasible routes and mission schedules for the ship. In the scenarios tested, the solution produced by the heuristic is not only generated in a fraction of the time of the current algorithms, but the total mission value collected is within 5% - 7% of those solution values., http://archive.org/details/aheuristiclgorit109453998, US Navy (USN) author., Approved for public release; distribution is unlimited.

Every two years, the Chief of Naval Operations is responsible for submitting the Program Objectives Memorandum to the Secretary of the Navy for further review and inclusion in the President's two-year budget input to Congress. The Chief of Naval Personnel's Strategic Resourcing Branch is challenged with building a manpower budget program that both meets the budget limitations set forth by Congress and the manning requirement choices made by Navy leadership. This thesis develops the Requirements-Driven Cost-Based Manpower Optimization (RCMOP) model. RCMOP is a linear optimization program designed to guide monthly values for officer inventory, promotions, accessions, designator transfers, and forced and natural losses. RCMOP's goal consists of minimizing a weighted penalty function of unmet manpower requirements while meeting the Navy's fiscal constraints over a two-year time horizon. Implementation of the test scenario shows that resulting costs fall within 10% of predicted budget estimates, and promotion metrics approximate the values expected by law and policy. The model also indicates a need to increase total OCS accessions (by 11%) with respected to projected values as well as the percentage of 1000-coded billets filled by staff and fleet support officers., http://archive.org/details/navyofficermanpo109454892, US Navy (USN) author., Approved for public release; distribution is unlimited.

This thesis considers the problem of protecting an electrical power grid against a potential attack on its physical infrastructure. We develop a mathematical model, called â Defense of Known Interdictionsâ (DKI), that identifies the optimal set of components to defend in an electrical power grid given limited defensive resources. For a small test network, we show that defending fewer than 10% of the buses reduces the possible disruption from an attack by over 20%. Previous research has developed optimization models, called I-DCOPF, to find optimal or near optimal interdiction plans for electrical power grids. DKI solution time is determined by I-DCOPF solution time. We develop a model, called the Network Dual Relaxation (NDR), to replace I-DCOPF and reduce solution times. NDR approximates electrical power grid behavior as a minimum cost network flow and uses this approximation to quickly estimate a lower bound for the exact interdiction model. We test NDR on a portion of the North American power grid with a computational limit of 6000 seconds. Results with ten buses defended show that NDR finds solutions that are, on average, 40% better than those of the exact I-DCOPF model with a significant reduction in computational time., http://archive.org/details/defendingelectri109453677, US Navy (USN) author., Approved for public release; distribution is unlimited.

Capital Investment Planning Aid (CIPA) is an optimization-based decision support system created for the U.S. Navy to help plan yearly force structure procurement and retirement. CIPA constraints include yearly industrial and budget limits, as well as mission inventory and force mix requirements. Over a 30-year planning horizon, CIPA helps plan over $1 trillion. Several approaches have been proposed and implemented to solve the CIPA core, a mixed-integer linear program (MILP). Unfortunately, some of these MILPs cannot be solved in a reasonable amount of time using general-purpose commercially available optimization software. This thesis presents a new MILP-based heuristic technique, fix-and-relax, that yields good quality solutions and reduces the computational solution time for our set of realistic test cases., http://archive.org/details/applyingfixandre109454325, Lieutenant Junior Grade, Turkish Navy, Approved for public release; distribution is unlimited.

This thesis evaluates the ability of the Xpress-MP software package to solve complex, iterative mathematicalprogramming problems. The impetus is the need to improve solution times for the VEGA software package, which identifies vulnerabilities to terrorist attacks in electric power grids. VEGA employs an iterative, optimizing heuristic, which may need to solve hundreds of related linear programs. This heuristic has been implemented in GAMS (General Algebraic Modeling System), whose inefficiencies in data handling and model generation mean that a modest, 50-iteration solution of a real-world problem can require over five hours to run. This slowness defeats VEGA's ultimate purpose, evaluating vulnerability-reducing structural improvements to a power grid. We demonstrate that Xpress-MP can reduce run times by 60%-85% because of its more efficient data handling, faster model generation, and the ability, lacking entirely in GAMS, to solve related models without regenerating each from scratch. Xpress-MP's modeling language, Mosel, encompasses a full-featured procedural language, also lacking in GAMS. This language enables a simpler, more modular and more maintainable implementation. We also demonstrate the value of VEGA's optimizing heuristic by comparing it to rule-based heuristics rules adapted from the literature. The optimizing heuristic is much more powerful., http://archive.org/details/anenhancedimplem109451996, Approved for public release; distribution is unlimited.

This thesis addresses the problem of optimal evacuation of a naval ship. We propose the use of a dynamic escape-route system which employs a signaling system to adapt the emergency egress process to the instigating contingency. The evacuation process is represented by a nonlinear network optimization model with an objective function that integrates two conflicting goals: the average evacuation time and the ship's integrity. The nonlinearity in the model results from (a) speed being a nonlinear function of concurrent flow on passageways, and (b) delays caused by opening closures. We also account for counter-flows and passageways used by repair parties. The problem is heuristically solved through an iterative process that updates speeds and delays as it proceeds, and dynamically adds valid inequalities to avoid counter-flows. A bound on the solution quality is obtained by solving the problem under optimistic conditions. Compared to static routes in a modern frigate, model solutions show that dynamic routes can improve the average evacuation time by 20%, reduce the time of the last evacuee by 25%, and improve ship integrity. We also demonstrate that even greater improvements are achievable with minor design changes in the ship., http://archive.org/details/dynamicescaperou109452085, Approved for public release; distribution is unlimited.

The Littoral Combat Ship, in its final steps toward employment, is an entirely new breed of U.S. Navy warship. USS Freedom (LCS 1), scheduled to be commissioned in May 2007, introduces an advanced technological platform. It includes but is not limited to several new optimal manning and training concepts such as SHIPTRAIN and SMARTSHIP introduced by the U.S. Navy. The Littoral Combat Ship Wholeness Concept of Operations requires a crew to certify in 15 mission areas using its core crew and one additional mission area applicable to both the core crew and mission module personnel. Using a discrete event simulation tool call the Total Crew Model, this study analyzed the currently proposed Fleet Response Training Plan for Littoral Combat Ship. An examination using a 14-day training cycle snapshot of the 40 propsed crewmembers was found to be sufficient to sustain the ship through a training assessment phase. The snapshot evaluated crew endurances using 63, 67, and 70-hour workweeks. The modeling showed the 70-hour work week satisfied the manpower requirement workload as delineated in OPNAVINST 1000.16J. This work week, however, exceeded core crew endurances by 594 hours and 42% of the crew exceeded applicable fatigue levels. The model's results indicate that eight additional core crewmembers are required to conduct the training assessment phase without exceeding core crew endurance., http://archive.org/details/alittoralcombats109453573, US Navy (USN) author., Approved for public release; distribution is unlimited.

Counterdrug operations are of national interest to the U.S. and our allies because the illegal production and trafficking of drugs threatens U.S. national security and undermines security and stability in Latin America. Since law enforcement tasked with counterdrug operations is not given enough platforms to search every location at all times, they must decide how to employ their scarce platforms. To assist law enforcement, we develop a defender-attacker optimization model that utilizes actionable intelligence to coordinate the simultaneous, cooperative disposition of law enforcement platforms in an optimal manner against a smuggler. The model utilizes stochastic dynamic programming to represent an intelligent smuggler, who has the ability to reevaluate his remaining path at decision points along his journey, based on knowledge obtained en route and expectations previously derived. The model employs Global Benders' Decomposition to determine the optimal placement of three different types of law enforcement platforms simultaneously prosecuting one of three possible types of smuggler. We show that such computations cannot be performed fast enough to be used in a tactical decision aid, since they typically require in excess of two hours. Upon further analysis using our model, we determine a large number of defender missions do not have a substantial impact on the attacker's risk. Based on the results of our model, we believe further algorithmic development is needed for implementation into a tactical decision aid to assist in counter drug operations., http://archive.org/details/optimalinterdict109455129, Outstanding Thesis, US Navy (USN) author, Approved for public release; distribution is unlimited.

Humanitarian assistance is of growing importance to the United States and the Department of Defense's strategic objectives. Thus, United States combatant commands increasingly rely on humanitarian assistance cargo transportation programs to deliver materiel to people in need in their areas of responsibility. This report analyzes the options available to these commands in seeking humanitarian assistance cargo transportation. The report offers a description of current operations, with a specific focus on the European area of responsibility, where these programs have had limited activity. The analysis reaches the following conclusions: (1) currently no transportation program exists that focuses on providing a quality of service to combatant commands and humanitarian assistance transportation needs, (2) legal, fiscal, and operational mechanisms exist and are outlined to create such a program, and (3) exclusively space-available transportation is generally insufficient for providing the quality of service that may be required for relationship-building through humanitarian assistance cargo transportation, and contract shipping may be necessary. These conclusions are placed in the context of current humanitarian assistance operations, and relevant operational considerations are highlighted throughout the report. The analysis is based on both a quantitative model of transportation, as well as detailed conversations with humanitarian assistance personnel throughout key Department of Defense organizations., http://archive.org/details/analysisofhumani109457334, Lieutenant, United States Navy

Wood (1995) to handle multiple types of interdiction assets (e.g., aircraft, ground-based inspection teams), referred to here as "inspectors." A single evader attempts to traverse a path between two vertices in a directed network while an interdictor, controlling inspectors of different types, attempts to detect the evader by assigning inspectors to edges in the network. Each edge has a known probability of detection if the evader traverses the edge when an inspector of a given type is present. The problem for the interdictor is to find a mixed inspector-to-edge assignment strategy that maximizes the average probability of detecting the evader, i.e., the "interdiction probability." The problem for the evader is to find a mixed "path-selection strategy" that minimizes the interdiction probability. The problem is formulated as a two-person zero-sum game with a surrogate objective that evaluates expected number of detections. That model is solved with a "direct solution procedure" and a "marginal-probability solution procedure." On numerous test problems, both procedures correctly compute expected number of detections, but the latter more often finds a solution that simultaneously optimizes interdiction probability. The latter procedure is also much faster and is therefore preferred., http://archive.org/details/twopersonzerosum109455334, Turkish Army author, Approved for public release; distribution is unlimited.

This thesis focuses on interdiction of Vehicle Borne Improvised Explosive Devices (VBIED) on a major city by using "transparent" and "deceptive" assets. Transparent assets (e.g., road blocks) are those for which we assume positions are known by both attackers and interdictors. "Decoys" and "traps" are deceptive assets. Decoys are meant to be perceived as effective interdiction assets by attackers, while traps are not perceived. We use a mathematical optimization model to allocate interdiction assets maximizing expected interdicted "value." Then, we use agent-based simulation to assess the effectiveness of those interdiction plans against a variety of attacker's behaviors: perceptive (as assumed by the optimization), naiÌ ve, communicative, route blocker (static), route blocker (dynamic) and clairvoyant. We use two test networks and seven scenarios consisting of different combinations of interdiction assets. From our analysis we note that: (a) if the network incorporates deception, any behavior other than perceptive may be advantageous to the attacker; (b) a communicative behavior proves effective for the attackers against scenarios containing traps; (c) decoys are most effective if used in defense against perceptive-like behaviors; and, (d) if the defender expects perceptive-like behavior, then adding transparent assets to traps and decoys may be of little value., http://archive.org/details/deceptivetactics109454207, US Navy (USN) author., Approved for public release; distribution is unlimited.

Theater Security Cooperation (TSC) is a U.S. strategy for improving multi-national relationships through cooperative efforts. Spitz develops the Central-West Africa Resource and Mission Allocation (CARMA) optimization model, which posits a naval vessel carrying various expeditionary partnership teams to transit an area of responsibility and conduct missions garnering the maximum amount of TSC value. CARMA can be solved with formal, mixed-integer optimization, at the expense of computational time. This thesis modifies the original Spitz's scenarios to test H-CARMA, a fast heuristic algorithm developed by Dwyer, and its performance under shorter planning horizons, multiple budget constraints and different distribution of missions and TSC value across countries. Most of the scenarios evidence shortcomings of H-CARMA that were not apparent in the earlier scenarios tested by Dwyer. In all but one of the reviewed cases, H-CARMA generates solutions with total TSC value less than 81% of those using Spitz's algorithms, and, in the worst of these cases, the solution only achieves 51 percent. When there is no slack in terms of time and budget, MIP solutions outperform those of H-CARMA by more than 25% in most cases examined. We identify sources for some of these deficiencies and recommend changes to address them., http://archive.org/details/assessmentofheur109454907, US Navy (USN) author., Approved for public release; distribution is unlimited.