|
|
|
|
MOR Journal Abstracts
|
 |
Number 1 |
|
Number 2 |
|
Number 3 |
|
Number 4 |
Grouping
Personnel for Performance Evaluation via Integer Programming (Lt Col Andrew
P. Armacost and Dr. James K. Lowe)
In their
2nd class (i.e., junior) year, cadets at the Air Force Academy are evaluated and
ranked by a board of senior officers for placement into career fields.
The cadets’ records are split into two groups before being scored.
Standardizing the scores of the two panels relies upon one key
assumption: that the two groups of cadets “look” the same.
To guarantee a fair split, Andy Armacost and Jim Lowe developed, tested,
and implemented an integer programming model that ensures equity of the groups
with respect to known, objective performance measures and key demographic
attributes. (Pg. 5)
Mission
Assignment and Maintenance Scheduling for Multi-State Systems (Dr. Thomas G.
Yeung, Dr. C. Richard
Cassady, P.E. and Dr. Edward A. Pohl)
In
today’s environment of shrinking budgets, most military organizations are
constantly faced with decisions on how to allocate these limited resources
effectively. This paper focuses on
the allocation of maintenance resources for a fleet of repairable systems.
Setting maintenance priorities among the different platforms and
different systems for a particular schedule is an extremely challenging problem.
In this paper, a model-based methodology is developed for managing the
maintenance strategies for systems that have multiple states (more than two).
A non-linear math model is developed for this problem.
Several heuristic’s are developed and analyzed.
A Monte-Carlo simulation is utilized to evaluate a variety of mission
profiles and system states. Our
methodology is applied for a hypothetical squadron of F-16 Aircraft although it
may be generalized to almost any military system containing a fleet of systems
the need to perform a set of missions. A
dispatching heuristic is developed that provides a practical means to obtain
quality solutions to such a complex combinatorial problem. (Pg. 19)
Midway Revisited: Detecting Deception by Analysis of Competing Hypotheses (Frank J. Stech and Christopher Elsaesser)
Historical
accounts of military deception abound, but there are few historical accounts of
counter-deception, and fewer operational theories. This paper describes a
business process and semi-automated tools for detecting deception. The
counter-deception business process begins with hypothesis generation. This
consists of automated course of action generation in tactical situations;
strategic situations require hypothesis elicitation from analysts. Next,
hypotheses and related potential evidence are represented by a Bayesian belief
network. This network is the basis of a diagnostic analysis derived from
classification theory. The result is a weighted list of possible observations
that: (1) identify distinguishing evidence that a deceiver must hide and a
counter-deceiver must uncover, (2) isolate local deception in intelligence
reporting and sensing from global deception, and (3) identify circumstances when
it might be fruitful to entertain additional hypotheses. We illustrate this
process by describing how it could have been used by the Imperial Japanese Navy
before the Battle of Midway to detect the American denial and deception tactics.
(Pg. 35)
A New
Sensor Allocation Algorithm for the Space Surveillance Network (Dr. James G.
(Gil) Miller)
The 1st Space Control Squadron has the mission to task the Space Surveillance Network to track satellites for US Strategic Command (USSTRATCOM) to maintain an accurate satellite catalog. An accurate and timely satellite catalog is fundamental to achieving Space Situation Awareness and supporting USSTRATCOM’s Space Control mission. Sensor tasking is a resource allocation problem, which is mathematically formulated as a constrained, non-linear integer programming problem. An operational prototype was developed using marginal analysis to solve this resource allocation problem. The operational prototype demonstrates improved performance against several key measures compared to the previous program, which employed a greedy algorithm. (Pg. 57)
A
Holistic Approach to U.S. Army Recruiting Input Allocation (MAJ John R.
Brence, Ph.D., LTC Michael J. Kwinn,
Jr., Ph.D. and Dr.
David A. Thomas)
Allocating
Army resources to meet recruiting mission requirements is a very sensitive and
important issue. When the military services meet their recruiting goals, the
analytical research focus is on how to keep recruiting constant while reducing
inputs such as recruiters and advertising.
Conversely, when the military services miss their goals, the analytical
research focus is on how to increase recruiting while keeping the inputs
constant. Today, we may be faced
with the challenge of expanding the military, requiring more recruits and
possibly less inputs. This research
analyzes the allocation of these inputs to improve recruiting policy and updates
the current recruiter allocation model to fit current needs. This paper was selected as Best in Working Group 20: Manpower
and Personnel MORSS 2004 and Finalist for 2004 Barchi Prize. (Pg. 5)
Bayesian
Stockpile Reliability Methodology for Complex Systems (Christine M.
Anderson-Cook, Todd Graves, Michael Hamada, Nicholas Hengartner, Valen E.
Johnson, C. Shane Reese and Alyson G. Wilson)
Logistics require that the U.S. Armed Forces to purchase weapons, ammunition and spare parts in large parts that are stored until the time they are used. In order to effectively understand the health of the stockpiles of these systems, it is important to have statistical models to estimate both current and future reliability at the component and system levels. This paper demonstrates how reliability can be monitored for an aging stockpile with a Bayesian analysis method for a complex system involving multiple components and sub-systems. (Pg. 25)
Reducing
Aircraft Down for Lack of Parts with Sporadic Demand (Dr. Tovey C. Bachman)
For sporadically demanded, critical spare parts, traditional methods of relating inventory investment to readiness fail. Setting stock levels for such items challenged operations researchers for over 30 years. In 1999, Tovey Bachman conceived a way of balancing the risks of being out of stock and over-investing in inventory for these parts—the Peak Policy. Supported by DLA’s Aging Aircraft Program, he developed this idea into a simulation-based method for setting stock levels, showing that it could reduce parts shortages and aircraft down as well. DLA began live testing in 2004 and wider implementation is underway. This paper was awarded the 2005 MORS Barchi Prize. (Pg. 39)
Two
Sensor Motion Models for the Dynamic Sensor Coverage Problem (1Lt Nathan P.
Yerrick, Dr.
Mesut Yavuz and
Dr. David E. Jeffcoat)
The
allocation of sensors is a significant problem in surveillance, search, and
tracking, with applications ranging from sea to space.
The widespread development of unmanned systems provides new opportunities
for these platforms to serve as intelligence, surveillance, and reconnaissance (ISR)
assets, pulling information from the battlefield into the global information
grid. This
paper incorporates ideas from automatic control, stochastic processes, and
optimization to solve a sensor coverage problem. The solution methods developed might be used to give
unmanned air vehicles the capability to make autonomous path planning decisions,
or to help a mission planner determine the number of target areas that could be
covered by a single ISR asset. (Pg. 55)
Optimal Placement of Suicide Bomber Detectors (Xiaofeng
Nie, Prof.
Rajan Batta, Prof. Colin Drury and Dr. Li Lin)
There have been an increasing number of suicide bombings on US interests, especially in the Middle East. This article addresses the tactical question of placement of SB detectors in a potential threat area. The objective is the minimization of the number of casualties. The sensitivity analysis allows a decision maker to select the appropriate number of detectors and the appropriate type of detector in the area being protected. The work builds off a recent paper by Kaplan and Kress in the Proceedings of the National Academy of Sciences. (Pg. 65)
Assessing situation awareness in networks of cooperating entities: A mathematical approach (Dr. Herman Monsuur)
In the field of Network Centric Warfare, for years people have been talking about the value of information, often without being able to quantify what they mean. This paper is a decisive step forward in this area. It presents a mathematical approach of situation awareness that is transferred between cooperating nodes in a network. The suggested method combines given quality characteristics of nodes, like information fusion centers, with network generated and updated values regarding situation awareness. A network measure of effectiveness determines the influence of network topologies; it can be used to evaluate investments in the (C4I) structure of the network. (Pg. 5)
The Impact of Network Centric Operations on Warfighter Effectiveness : A Study Using Agent-Based Models (Isaac Porche, Bradley Wilson and Susan Witty)
RAND Arroyo Center was tasked to “quantify the impact of cognitive capability on the performance of network-centric operations” by the Chief Information Officer (G6) of the U.S. Army and the Office of the Assistant Secretary of the Army for Acquisition Logistics and Technology. Using agent-based force-on-force simulation experiments, Isaac Porche, Bradley Wilson, and Susan Witty addressed the relative impact of networking parameters as well as the importance of force makeup on overall warfighter effectiveness. An abridged version of this report was presented at the International Control and Command Research and Technology Symposium in June 2005. (Pg. 17)
A Theory of Wide Area Search and Engagement (Lt Col Doug Decker, David R. Jacques and Meir Pachter)
A system analysis of wide area autonomous search and attack munitions is performed. The developed model captures the important parameters of target density in the battle space, the presence and density of false targets in the battle space, the performance of the seeker and of the Autonomous Target Recognition (ATR) module, air vehicle performance data, e.g., speed and seeker swath width, and munitions parameters such as warhead lethality. Applied probability theory is used to analytically derive the probability of success in search and engagement. operations. It is shown how the analytically derived results can be applied to the design and operation of an autonomous wide area search and attack weapon system. (Pg. 37)
A Tabu Search Approach to the Strategic Airlift Problem (COL Garrett R. Lambert, J. Wesley Barnes, Lt Col David A. Van Veldhuizen)
Headquarters Air Mobility Command (AMC) currently uses descriptive tools to analyze Strategic Airlift Problems (SAPs). These tools provide neither optimized flight schedules nor operational-level detail.
For the last 8 years, a Consortium funded by AFOSR and composed of personnel from AMC, the University of Texas at Austin, and the Air Force Institute of Technology, has developed cutting-edge solution methodologies for high-priority issues associated with end-to-end mobility problems like the SAP. This paper documents an effective and efficient SAP solution methodology, a robust dynamic adaptive Tabu Search (TS) algorithm, SAP-TS, which can provide AMC planners with sufficient operational-level detail to effectively and efficiently construct detailed routing and scheduling of strategic airlift resources.
An earlier version of this paper was invited to compete in the 72nd MORS Symposium Barchi Prize Competition. (Pg. 59)
Forecasting Israeli-Palestinian Conflict with Hidden Markov Models (LTC Robert Shearer)
The genesis of the Israeli-Palestinian conflict dates back to Arab protests of Jewish setters in 1891 and continues to the present time. This paper applies pattern classification techniques to the conflict in order to capture the patterns of escalation and forecast future escalations. These techniques provide insight into the dynamic behavior of the conflict and describe potential futures of the conflict in probabilistic terms, to include the likelihood of conflict, the time to conflict, and the time in conflict. (Pg. 5)
Integrating Bayes’ Theorem with Dynamic Programming for Optimal Intelligence Collection (Prof Yacov Y. Haimes, Zhenyu Yan and Barry M. Horowitz)
Daily, security agencies receive a plethora of data, information, and other intelligence reports on threats to the homeland. There is a crucial need for quantitative and systemic intelligence analyses. To perform Bayesian analysis of intelligence, an important task is to collect intelligence. Within a given budget, one must determine a feasible path of scenario tracking, intelligence gathering, and analysis which maximizes the effectiveness of intelligence collection. Clearly, the solution efforts to this combinatorial problem could become prohibitive for large networks with many sub-scenarios. Yacov Haimes, Zhenyu Yan and Barry Horowitz developed an approach for addressing intelligence collection by integrating Bayesian analysis with dynamic programming . With this approach, one can determine an optimal path (or paths) of intelligence collection in each stage within a given budget.(Pg. 17)
The Fast Theater Model (FATHM) (Gerald G. Brown and Alan R. Washburn)
FATHM is a two-sided theater model of air-ground warfare, designed to emphasize the effect of platform counts and munitions logistics on battle results. The paper first explains FATHM’s genesis as a fusion of the Army’s COSAGE model and the Air Force’s CFAM model, including some experiences of the authors in fusing two models that were not originally designed to be fused. FATHM’s imitation of COSAGE is through a Lanchester system that obtains its coefficients from the COSAGE killer-victim scoreboards. The ensuing ordinary differential equations are solved numerically in a sequential scheme that permits regular air strikes in accordance with CFAM. Each of these air strikes requires the solution of an embedded linear program. The war proceeds in phases, with each phase having its own COSAGE killer-victim scoreboard. Attrition to both sides is accumulated over a war that may last for months, and finally output for study by the user. The entire process of fighting the war requires only a few minutes, thus justifying the “fast” part of FATHM’s name. This makes FATHM particularly useful when many different scenarios must be compared. (Pg. 33)
Surveying The Spectrum Of Human Behaviour In Front Line Combat (D. Rowland and L. R. Speight)
Combat models, if they are based on performance estimates obtained in peacetime conditions, can yield predictions of attrition rates that greatly exceed those actually observed in historical battles. Previous analysis by the first author has shown that achieved kill rates in realistic field trials can be a factor of ten times lower than those obtained on the firing range. The rates actually achieved in historical battles are typically ten times lower than those seen in field trials. The key to this last effect is the impact that live combat can have on the behaviour of trained soldiers in the field. The heroic few are relatively unaffected, and indeed their willingness to expose themselves to danger in order to further the unit’s mission can be an inspiration to their comrades-in-arms. However, for the remainder the range of behaviour can stretch from competence, but without initiative, right through to non-participation. Obtaining accurate descriptions and quantitative estimates of this range of behaviours is fraught with difficulty, not least because official histories, for obvious reasons, tend to skate over those instances where military performance falls some way short of the ideal. This paper draws on a number of sources in order to make good this deficiency. Gallantry awards have been studied in order to define the heroic end of the spectrum, and battle records and disciplinary records to define that of non-participation. Detailed accounts of individual engagements have been utilised in order to describe and quantify intermediate levels of performance, relying on the recorded observations of those directly involved, mainly at lower levels of command (junior officers and NCOs). The results so obtained have been compared with those yielded by other studies of combat behaviour. These other studies have employed different methodologies and have utilised different performance criteria. Nevertheless, their findings seem to be both mutually consistent and broadly in accord with our own estimates. (Pg. 47)
Military Operations Research Society (MORS) Oral History Project Interview of Hork Dimon, FS (Bob Sheldon, FS)
No Executive Summary (Pg. 61)
Some
of our links are in portable document format (pdf); they will be indicated by
this symbol
|
