|
|
MORS seeks to recognize excellence in the military operations research profession. A person who best represents the ideals of and has made significant contributions to the operations research community and MORS is selected annually to receive the Vance R. Warmer Award, presented since 1978 in honor of the first Executive Secretary of the Society. The Rist Prize, named after David Rist (an early Director) and first presented in 1965, is awarded annually for the best paper submitted in response to a call for papers. The Barchi Prize, presented since 1983 in honor of CDR Richard H. Barchi (a former Director) recognizes the best paper presented at the annual Symposium. The Graduate Research Prize is presented for the best thesis in each operations research class at the Air Force Institute of Technology and Naval Postgraduate School. Rist Prize1998 - "The Generation, Use, and Misuse Of "PKs" in Vulnerability/Lethality Analyses," Paul H. Deitz and Michael W. StarksAbstract: Beginning with World War II and its aftermath, the area of ballistic vulnerability/lethality (V/L) was first defined as a specific discipline within the field of ballistics. As the field developed, various practices and metrics emerged. In some cases metrics were developed that were abstractly useful but bore no direct relationship to field observables. In the last decade, as issues concerning Live-Fire strategies have gained importance, increased attention has been focused on V/L with the intent of bringing greater rigor and clarity tot the discipline. In part this effort has taken the form of defining a V/L Taxonomy which is a method of decomposing a series of concatenated complex processes into separable, less-complex operations, each with certain specifiable properties and relationships. Using the Taxonomy, this paper describes how the most commonly used V/L metrics are a function of platform aggregate damage, reduced platform capability; and reduced platform military utility. We show that these three distinct and separable classes of metrics are linked by operators that are multivariate, stochastic, and nonlinear. We also show that it is useful to form probability distributions with respect to initial and boundary conditions in order to characterize damage, capability, and utility. Many defense community studies ignore these distinctions to the detriment of fundamental clarity. Examples are given and potential remedies described. 1999 - "Signals from Space: The Next-Generation Global Positioning System," Lee J. Lehmkuhl, David J. Lucia and James K. FeldmanAbstract: The Global Positioning System (GPS) is a constellation of satellites that provides precise navigation and timing information to military and civilian users worldwide. GPS signals from space guide cruise missiles and rental cars, and allow us to track the locations of railroad boxcars, golf carts, and soldiers in the field. As the provider of this national and international asset, the US has a vested interest in seeing that GPS remains the premier space-based navigation system, and has embarked on a GPS modernization program. Improvements in signal generation and processing technology now allow us to consider new signal structures, which will greatly improve the usefulness of GPS for military and civilian users. Choosing between these new signals, however, presents senior decision makers with a host of both technical and operational tradeoffs, many between competing military and civilian interests. The decision analysis presented here modeled the value of GPS to different user communities and quantified the tradeoffs. The results allowed the GPS Independent Review Team to recommend a new signal with superior military value that also meets all civilian technical performance requirements. 2000 - "Why Skill Matters in Combat Outcomes: and How to Include It in Combat Modeling," Michael Fischerkeller, Wade Hinkle and Stephen D. BiddleAbstract: Combat assessment and force balance methodologies will play important roles during the next Quadrennial Defense Review in analyzing the capabilities of postulated forces, in planning scenarios, and in determining the proper balance between readiness and modernization. Research at IDA suggests that the analytic tools currently used for these purposes may substantially undervalue the contributions of military skill and advanced operational concepts. Our work in this area has won two awards: the MORS 1997 Barchi Prize and a 1999 MORSS Medal for Excellence in Operations Research. This project used a combination of statistical analysis of historical data, combat simulation experimentation, and close examination of critical historical cases to develop and initially test a formal set of hypotheses about how technology, skill, and operational concepts interact to produce combat outcomes. This presentation to the MORS community will offer a summary of the research-to-date and a discussion of whether and how the resulting mathematical model ought to be included in QDR-related efforts to improve existing analytic tools. 2001 - "The Development of an Information-Based Direct Fire Attrition Structure in AWARS," H. Kent Pickett and W Peter CherryAbstract: Over the past 30 years, the Army has relied on a Lanchester-based attrition structure in its aggregate level combat models. This structure, commonly called the Bonder/Farrell Attrition Algorithm, is based on parameters describing weapon performance (probability of kill given a shot and probability of detection given target exposure), target exposure, and firer preference for particular targets. By the mid-1990s, it became apparent that the Napoleonic process, represented by combat simulations of marching units together and fighting them until one broke and ran, was not the battle Army commanders expected to fight in the near term or the future. This paper describes a new approach to simulating the direct fire process by the Army Warfighting Simulation (AWARS). The AWARS methodology uses the Bonder/Farrell attrition rate process but attempts to represent other aspects of the battle affecting internal unit coherence – in particular, the state of information a unit has about other friendly/enemy units and its own situation. The paper contains a description of the overall architecture of the AWARS model affecting the direct fire process, unit geometry, the representation of key parameters in orders, situation maps, and the timing structure driving the simulation. Finally, a set of succeeding states is described allowing units to establish successively higher levels of unit command, control, and a coherent fire mechanism (providing more accurate fields of fire and a higher knowledge of enemy location and intentions). The paper also describes the ability of the methodology to represent battle “lulls” (these are the times when individual vehicle crews simply do not engage targets at their most efficient rate). 2003 - "Input-Output Modeling for Effects-Based Operations,"
Abstract:
Leontief
developed the input-output model to make macro-economic assessments based on the interdependencies of various production sectors
within a region. We discuss the concept of Effects-Based Operations (EBO) and
its inherent requirement for analytical modeling to assess the effects of
potential actions. We propose that
the input-output model may be used to assess the direct and indirect impact of
military operations in an enemy country. We
present the input-output model and demonstrate how it can be used to assess the
impacts of a variety of military strategies against a region or nation.
|
 |
Brigade
Combat Teams (BCTs). We analyzed the number of BCTs the Army needs to
meet operational requirements and estimated the stress on the force at
various levels of commitment.
This analysis covered both current and future demand scenarios with
corresponding force structure options.
Our conclusions advised HQDA regarding force sufficiency and force
stress in both the near-term and future cases.
These conclusions also included assessments of alternative force
structure options.
In particular, we determined the circumstance under which the Army
could “run out of BCTs” and how frequently units can expect to deploy. |
|
Support
Structure. We analyzed how well the planned Army force structure will
meet operation requirements in potential demand scenarios.
This analysis included all deployable Army units across the Active
and Reserve Components.
We identified which types of units are unable to meet operational
commitments as well as which types of units are likely to be over or under
stressed. We
also identified which unit types have improper AC/RC balance and which types
of units should and should not be managed under ARFORGEN. |
|
Sustain-Surge.
We analyzed a variety of scenarios that combined different levels of
sustained, steady-state operation and surge operations.
For example, if the Army is maintaining a sustained commitment of X
BCTs, it has the capacity to surge with Y addition BCTs.
As the level of sustained commitment increases, rotational stress
will increase accordingly while surge capability decreases.
This analysis identified the frontiers between various commitment
levels, stress thresholds, and force sufficiency. |
|
ARFORGEN.
We developed MARATHON to replicate ARFORGEN in accordance with
emerging concepts from HQDA G-3 and Forces Command (FORSCOM).
Because we created the model simultaneously with the evolution of the
ARFORGEN concept, we were able to directly impact ARFORGEN development by
G-3 and implementation by FORSCOM.
In particular, we assessed that if ARFORGEN is implemented with
flexible lifecycles, it will save the Army 2-4 BCTs of AC force structure.
We also assessed that the Army can maintain the greatest number of
available units by evenly distributing capabilities over time.
Both of these conclusions were accepted by HQDA for ARFORGEN
implementation. |
|
Access
to the Reserve Component.
We employed MARATHON to examine access to the Reserve Component.
This analysis showed how various policies regarding RC access impact
Army force structure, force sufficiency, and deployment tempos.
In particular, it showed which policies are vest for different types
of units and operation conditions.
ASA(M&RA) accepted the results of this research and directed
dissemination to HQDA (G-1, G-3, and G-8), FORSCOM, NGB, and USAR. |
Our
methodology will continue to have far-reaching impacts on the Army:
|
We
developed a personnel extension to the model in support of HQDA G-1 and
Human Resources Command.
This model extension allows us to examine various personnel policies
under ARFORGEN by simulating the movement of soldiers through their careers,
to include assignment to units that are moving through ARFORGEN operational
readiness cycles.
The model shows how personnel policies affect unit fill-rates and
soldiers’ availability for schools and assignments. |
|
We
are developing an extension to the model to analyze equipping issue for HQDA
G-8. This
model allows us to examine assignment policies for training equipment,
deployment equipment, and pre-positioned stocks of equipment.
It also allows us to examine the affects of cyclic readiness and
deployments on decisions to modernize, replace or recapitalize equipment. |
Summary
of Implementation:
|
Development
of the MARATHON stimulation model; briefed at G-8 OPD (Jan 05) and used for
all of CAAs ARFORGEN-related analyses (Jan 05 – present). |
|
Analysis
of ARFORGEN concept; used by HQDA G-3 for ARFORGEN approval (Jun 05); used
by FORSCOM for ARFORGEN implementation (Apr 05 – Oct 05) |
|
Analysis
of force structure (BCTs as well as CS/CSS structure); used for QDR and
Operational Availability 06 (Mar – Sep 05) |
|
Analysis
of BCT force structure in the near term; used by QDA (Aug 05) and
Heavy-Infantry mix decision (Oct 05) |
|
Analysis
of sustained vs. surge operations (BCTs as well as CS/CSS structure; used
for QDR (Aug-Oct 05) |
|
Analysis
of access to the Reserve Component; used for QDR (Oct 05) |
|
Development
of the MARATHON-PER simulation; deliver to HQDA G-1 for personnel analysis
(May 05) |
|
Development of the MARATHON-EQUIP simulation; used to analyze the Equipment Maneuver Plan for HQDA G-8 for ACP DP41 (working) |
2007- “Countering Radio Controlled Improvised Explosive Devices,” Dr. Edward S. Michlovich
Showing the effectiveness of an operation designed to prevent something from happening has always been a challenge. The inability to prove an effect can cast doubt on the value of the operation and lead to a classic question: When is an uncertain gain worth the known costs associated with it? In mid-2004, operational commanders were facing such a situation. In this case, the question was magnified by the exigent circumstances of war, where lives were quite directly in the balance.
An operation intended to help protect soldiers and marines from radio-controlled IED (RCIED) detonations was being conducted in Iraq. However, there were serious concerns about the operation’s cost, particularly the ramifications of high utilization on certain assets. In the absence of evidence that the operation was effective, some consideration was given to curtailing or terminating it. To inform decisions on the future of the operation, the Marine Corps asked if I could find a way to assess effectiveness.
I gathered data from units in theater and devised a technique, based on bootstrap statistical theory, to pull out meaningful results. The methodology is conceptually straightforward, consisting of three fundamental steps: calculate a metric relevant to effectiveness of the operation (IED frequency, for example); develop the distribution of values that the metric would take if the operation were not effective; compare the latter to the former. This becomes the equivalent of a statistical significance test, with the proposition that the asset had no effect being the null hypothesis.
The most challenging part of the methodology is the second step. This is where bootstrap theory comes in—it allowed me to use the data from when the asset was not employed to generate estimates of IED frequency if the asset had not been effective. Repeating the calculation many times in a Monte Carlo-like fashion allowed me to generate the necessary statistical distribution.
Analysis of the initial months of the operation revealed a correlation between the use of the asset and a marked reduction in IED frequency. I found a high statistical confidence (>95%) associated with the correlation; in statistical terms, I could reject the null hypothesis of no-effect. The analysis further showed that the operation’s effectiveness was limited only by the paucity of assets dedicated to it.
In early 2005, the Director, Joint IED Defeat Task Force, hosted my presentation of the analysis to Commander, U.S. Central Command, GEN John Abizaid. As a result of the analysis, he requested additional assets be devoted to the operation. Indeed, since the initial analysis, use of the asset has tripled. Follow-on analysis verified the original findings — the operation continued to be effective in countering RCIEDs. Calculations indicate that the continued and expanded use of the asset — which this analysis is generally credited with — has likely resulted in the prevention of hundreds of casualties in Iraq. CNA has since expanded the analysis to related operations as well.
Abstract: Over the last several years, a new approach to model-based analysis has been developed at RAND. This approach, exploratory analysis, greatly expands on traditional analytic approaches in order to enhance understanding of complex problems, provide a wider range of information for decision makers, improve comparison between alternative models, and thereby enable greater comprehension of policy options. This paper reviews the methodology of exploratory analysis and its advantages over traditional analysis in the context of a search for the preferred weapon mix. We begin by walking through a traditional analytic approach and showing the kinds of results that are often observed. We then perform exploratory analysis, requiring a large number of computational experiments-on the same problem, and show that it provides more information and keener insights than we originally obtained. We continue by describing exploratory analysis more generally, and demonstrate its benefits to the decision maker and the analyst. We also discuss what is required for its routine use.
Abstract: The engineering of complex systems of systems has received greatly increased attention in recent years. Although the characteristics and system engineering challenges associated with systems of systems are well understood, effective architecting approaches that enable cost/performance trades are still immature.
A systematic approach to considering how best to upgrade specific, complex systems of systems is postulated and demonstrated. Treating cost as the independent variable (CAIV), it seeks to find the "best" point design that may involve upgrading all component systems simultaneously, not just one at a time. The process has been demonstrated on a naval mine countermeasures system of systems representation of sufficient complexity to demonstrate feasibility of the approach. A constrained, nonlinear optimization problem is formulated whose objective function is a representation of the top-level measure of effectiveness (MOE), with constraints represented by functionalized Performance-Based Cost Models, secondary MOEs, and technology-driven bounds on system measures of performance (MOPs). Both closed-form and simulation-based optimization approaches have been demonstrated, including an efficient constrained stochastic optimization method necessitated by the use of simulation to generate MOEs.
This quantitative process for developing system of systems upgrade options for very complex situations can result in more effective and comprehensive systems acquisition and technology investment strategies.
Abstract: In this paper we consider the dynamic routing of unmanned aerial vehicles (UAVs) currently in operational use with the US Air Force. Dynamic vehicle routing problems (VRP) have always been challenging, and the airborne version of the VRP adds dimensions and difficulties not present in typical ground-based applications. Previous UAV routing work has focused primarily on static, pre-planned situations; however, scheduling military operations, which are often ad-hoc, drives the need for a dynamic route solver that can respond to rapidly evolving problem constraints. With these considerations in mind, we examine the use of a Java-encoded metaheuristic to solve these dynamic routing problems, explore its operation with several general problem classes, and look at the advantages it provides in sample UAV routing problems. The end routine provides routing information for a UAV virtual battlespace simulation and allows dynamic routing of operational missions.
Abstract: Military planning and analysis of operations on the Korean Peninsula require understanding of the potential North Korean (nK) ground attack. Understanding how these factors interact is essential in developing many products required by both planners and analysts. Unfortunately, many tools currently utilized to gain such an understanding are deterministic, thus failing to take into account the variability inherent in the operation, and time intensive, discouraging sensitivity analysis. The Center for Amy Analysis has developed a network model, the Troop and Equipment Movement Model that provides a stochastic analytical tool to help in understanding how such an attack could occur. The model simulates the movement of North Korean forces from their current garrisons to their respective mobility corridors along the Korean demilitarized zone. As output, the model provides numerous quantitative reports that support joint Army/Air Force decision making processes.
Abstract:
Air-to-air combat often is characterized by tradeoff decisions
involving lethality and survivability. For example, the choices of radar mode
and shooter maneuvers that are best for the success of a missile (lethality)
place the shooter at greater risk to a successful engagement by his adversary
(survivability). This paper documents the results of an operational test and
evaluation (OT&E) assessing these tradeoffs using operations research tools.
The test had two objectives: (1) assess the lethality of the AIM-120 air-to-air
missile and the validity of F-15-computed envelopes and displays as a function
of radar mode and shooter post-launch maneuvers, and (2) evaluate the effects of
pot-launch maneuvering on survivability during air-to-air engagements. There
were three noteworthy aspects of this evaluation. The use of Design of
Experiment (DOE) and statistical techniques to plan the test, execute flight
missions and simulations, and analyze results. The second was the marriage of
open-air flight test and simulation models to obtain realistic and tactically
sound test data. The third was a set of conclusions and recommendations
pertaining to the tactical decisions (tradeoffs) regarding radar mode and
shooter maneuvers, and the impact of those decisions on engagement success.
Abstract:
Military
planners, defense analysts, diplomats, and legislators need better tools and
models to provide them with better insights into where, when and to what extent country instabilities
might challenge national security interests so they can anticipate, plan and
budget for these possible occurrences in advance. This study draws upon the
state strength literature, uses recently developed data-mining tools and draws
upon an extensive database that includes annually aggregated data covering
political, economic, and socio-cultural domains for some 159 countries over the
period 1975-1999 to forecast the likelihood that countries throughout the world
will experience a certain level of intensity of instability over the period
2001-2015. The study uses a pattern classification algorithm-Fuzzy Analysis of
Statistical Evidence (FASE)—developed by Chen (2000) on behalf of the U.S.
Army to identify and analyze the relationships between country macro-structural
factors and historical occurrences of instability. A split-sample validation
design is used to evaluate the ability of FASE to generate competent
predictions, using the standard performance metrics overall accuracy, recall and
precision. The results demonstrate the potential capability of the model to
accurately forecast not just the occurrence, but also the level of intensity of
country instability six years in advance with about 80% overall accuracy. The
forecasts generated through the year 2015 suggest that South Asia and East
Africa will continue to harbor highly unstable states. However, most of the
states expected to improve their prospects for greater stability are also
located in these regions.
2004 - "Unit Manning," Mike McGinnis and Dave Saunders
Abstract: This paper outlines work of the authors in support of the Unit Manning initiative directed by the Secretary of the Army. In this work manning concepts are explained and analyzed, Unit Manning concepts are developed, and a scheduling model that was utilized to analyze the effect of Unit Manning on the Army as a whole is discussed. The heuristic model is rules based and calculates a Unit Manning and Unit Rotation Schedule of the Army’s 33 Brigade Combat Teams. The model conducts two passes over the time horizon, first calculating a rotation schedule and then overlaying a unit manning schedule on those rotations. The rules are adjustable by fourteen inputs, thus allowing the user to determine the effects of policies throughout the 164-month planning horizon. Output of the model consists of a Manning and Rotation schedule that additionally supports analysis of transformation decisions, and identification of friction points in terms of unit availability and personnel requirements.
2005
- “Reducing Aircraft Down
for Lack of Parts with Sporadic Demand,” Tovey
C. Bachman
Abstract: In the military aerospace environment, certain repair parts are infrequently demanded, but stocked because they are essential to maintaining a weapon system critical to the war-fighter. Because of their sporadic demand, it is difficult to decide when to buy these items and in what quantities. As systems become more reliable and failure rates decrease, the number of these infrequently demanded parts is likely to grow. Earlier studies for the Defense Logistics Agency (DLA) and the Federal Aviation Administration (FAA)—organizations that manage parts inventories for repairing complex systems—found the Peak ordering policy the author invented significantly reduced wholesale wait-time and backorders. Rigorous new experiments confirm the benefits of the Peak policy, and show it can reduce retail wait-time and backorders as well. By considering the distribution of retail backorders, or “holes,” over an aircraft squadron, we estimate the resulting reduction in the number of aircraft down for lack of parts. We also analyze the policy’s near-term effect on inventory value and procurement workload, showing that the Peak policy can reduce both within a few years of policy initiation. After 5 years of development and review, the Peak policy is mature enough for implementation. A live test is underway, and broader implementation is under consideration.
2006 - "Strategic Mobility: Building a Stochastic Critical Path Model," Stephen Nolan and Robert Stevenson
Abstract:
In 2004 the Mission Area Analysis (MAA) Lab, Marine Corps Combat Development
Center (MCCDC) was asked the question “What does it take to move the 2015
Marine Expeditionary Brigade (MEB) to the Strait of Hormuz in 14 Days?”
A series of models of increasing scope and complexity were built to
answer this question, culminating in “Project Arnold”, a stochastic Critical
Path Model (CPM) with 5 parametric and 8 stochastic variables.
Project Arnold resulted in over 690,000 model runs, from which critical
insights are being drawn for the Marine Corps leadership.
This
paper describes the “spiral” development of the series of models from
brainstorming sketches on a white board through an Excel-based CPM, integration
of Microsoft Access, the decision to move to stochastic variables, and the use
of multiple PCs to process the data.
With each loop in the spiral, more questions arose yielding greater
scope, which in turn demanded greater capability from the model and innovation
from the analysts.
The thoughts, problems and work-arounds through each cycle in the process
are the focus of this paper, with results and insights included.
The Air Tasking and Efficiency Model (ATEM) plans intra-theater airlift of passengers and cargo pallets. United States Central Command (CENTCOM) provides theater-wide logistics support, including airlift, for Operation Iraqi Freedom (OIF), Operation Enduring Freedom (OEF) in Afghanistan, and other operations in surrounding areas and the Horn of Africa. For OIF, CENTCOM decides daily how to use about 35 cargo airplanes of up to five different types, based at five different airfields, to move passengers and cargo pallets between many origin-destination pairs from a set of about 20 airfields. Daily, each airplane and aircrew should fly no more than a maximum number of hours, must refuel within a limited number of hours at an airfield where fuel is available, and can make no more than a maximum number of landings overall and no more than a maximum number of landings “in the box” (i.e., in a combat area). ATEM plans this intra-theater airlift by optimally selecting for each airplane a configuration — a layout of seats accommodating a number of passengers and of pallets — and assigning a route that, in concert with all other airplane configurations and routes, moves as many passengers and pallets as possible between origin and destination airfields. CENTCOM planners have used ATEM to design monthly channel schedules for OIF and OEF, choose airplane bed-down airfields, optimize fleet mix of airplane types, determine where to fuel airplanes, and schedule daily catch-up operations. ATEM has helped move more with fewer airplanes.
"That I have hoisted sail to all the winds, which should transport me farthest from your sight." Shakespeare, Sonnett
|
Some
of our links are in portable document format (pdf); they will be indicated by
this symbol
|
