Track 1: Military Applications I: Modeling and Analysis
What aspects of O.R., advanced analytics, artificial intelligence, or machine learning are missing or underutilized within the military O.R. community? How can their incorporation better increase the rigorous and impactfulness of analysis provided to military decision-makers?
-
Many people think of processes as intentional groups of activities which serve profit maximizing businesses or mission driven government agencies. While this is true, it is an incomplete picture. Organizational process ecosystems exist which are emergent and self-organizing with no human intention. Dynamic human organizations reside within complex ecosystems where they both affect and are affected by their exogenous environments. Exogenous information stimulates critical infrastructure organizations including corporations and military units. May organizational reactions to exogenous stimuli be understood empirically and probabilistically? Are organizational reactions generalizable across organizations and cultures? In this discussion, we explore these questions and relate them to national security risks unique to the Information Age. Results from a recent study which modeled an emergent corporate communication ecosystem from semi-structured email data using explainable AI are presented along with a future work agenda.
John Bicknell
John Bicknell founded More Cowbell Unlimited to help America remain a beacon of hope and strength on the world stage. America must adopt Process Dominance as a core capability in order to innovate and survive in the Information Age. John is a national security thought leader and passionate analytics visionary. He has written extensively on national security matters related to information warfare, critical infrastructure defense, and space situational awareness.
-
Considered are resource allocation, patrolling, and hide and seek games between a defender and various types of adversaries. If the defender does not know the adversary’s type, then a Bayesian game arises. Whenever it is possible, conditions for the existence of a unique equilibrium are derived, and both equilibria and value functions are obtained in closed form. In some cases, the analyses show that there exists a cut-off index determining the set of targets that will be covered in the equilibrium strategies. Extensions to multiple defender and multiple adversary teams are also studied. Computational examples are provided.
Melike Baykal-Gursoy
-
Measuring operational readiness is expensive in terms of financial costs, time, and risk, and is usually perishable. While improved readiness accounting across defense and national security communities will likely be an evolutionary process with inputs from numerous stakeholders, Operations Research analysts are well suited to recommend some key features for an improved readiness representation framework that is useful, logically consistent, and most importantly is simple enough for incremental adoption by leaders at all levels. In this talk, we provide working prototypes of models to demonstrate the underlying principles associated with this new readiness representation approach.
Shaun Doheney
Shaun Doheney is the Chief Analytics Officer for JDSAT – a certified Service-Disabled Veteran-Owned Small Business specializing in Operations Research and Data Science. He is also the Chair of Resources and Readiness Applications at ProbabilityManagement.org – a nonprofit devoted to making uncertainty actionable. He is an INFORMS Certified Analytics Professional (CAP®) and PMI-certified Project Management Professional (PMP®) with a B.S. in Mathematics, an M.S. in Operations Analysis, and a Graduate Certificate in Data Analytics. As a Marine Corps Lieutenant Colonel (Retired) and Marine Operations Research Analyst, he has performed quantitative and qualitative analyses and evaluations across major DoD decision support processes.
-
Closed form approximations of supply chain metrics often rely on assumptions that are not well aligned with real world environments. Discrete Event Simulation (DES) offers an attractive alternative. With the Wholesale Inventory Model for Enterprise Resource Planning (WIOM-ERP) we introduce a modeling architecture in which a wide decision space of candidate inventory policies are evaluated in a DES and then assessed by an open form optimization model. Supporting the core WIOM-ERP components we build a modular tool chain, which allows the user flexibility of engagement, while automating tasks such as data upload, results archiving, post run analysis, and results export.
Duncan Robert Ellis
CDR Duncan Ellis is the Director of Supply Chain Operations Research for Naval Supply Systems Command in Mechanicsburg, PA. He holds a master’s degree in Operations Research from Naval Postgraduate School and a bachelor’s degree in Computer Science from University of Iowa.
-
Coverage is one of the most important performance metrics for sensor networks. In this study, we consider a hybrid area, point and barrier coverage application in which the decision-maker seeks to locate a number of sensors along a two-dimensional belt-shaped region. The sensors are assumed to have different characteristics and performances against different types of targets. We first formulate two mathematical models for the location problem and then develop a heuristic algorithm.
Mumtaz Karatas
Cdr. Mumtaz Karatas is an associate professor in the Department of Industrial Engineering at the National Defense University, Turkish Naval Academy. His work focuses on logistics and location planning problems in security applications. This work was supported by the Scientific and Technological Research Council of Turkey (TÜBITAK) with Grant #: 118E694.
-
We consider the problem of routing a UAV for the purpose of information gathering in an area where targets that contain valuable information emerge in time and space according to non-homogeneous Poisson process. The goal is to find a time-constrained route for the UAV while seeking to maximize the expected information gained. Numerical results and theory related to this problem will be presented. Sensitivity analysis with respect to arrival rates and their disparity will also be presented. Further, extensions to the case of multiple UAVs will be discussed.
Pin-Chun Cho
-
One major problem in air defense is to identify how to best position weapon systems. Lockheed Martin used a discrete definition of an objective function and a set of constraints. Our objective is to advance Lockheed Martin’s problem statement using stochastic differential equations and derive a metaheuristic algorithm which streamlines solution sets to provide recommendations for improvement of defense effectiveness.
Nikita Pereverzin
Nikita Pereverzin is a West Point cadet and soon to be Commissioned Air Defense Officer in the US Army. He is majoring in Mathematical Sciences with a focus in Computer Science.
Track 2: Military Applications II: Holistic Applications
-
One of the most important aspects for an analytics project is the ‘last mile’ of delivery. This talk will cover several projects that the Center for Strategic and Budgetary Assessments (CSBA) has performed for stakeholders at the Secretary and Congressional levels. Specific topics we will discuss include the Air Force Inventory Study, Studies in Logistics and the Defense Industrial Base, and work at the frontiers of AI and wargaming.
The common thread through all of these studies is translating the stakeholder’s question into a problem amenable for analysis, The key feature of this presentation will be in how the analytic results are translated and formatted to a form that impact policy.
Bryan Clark
Bryan Clark is a Senior Fellow at the Center for Strategic and Budgetary Assessments. At CSBA he has led studies in naval warfare, electromagnetic warfare, precision strike, and air defense. A career U.S. Navy submariner, his last position was Special Assistant to the Chief of Naval Operations and Director of his Commander’s Action Group, where he helped develop Navy strategy and implement new initiatives in electromagnetic spectrum operations, undersea warfare, expeditionary operations and personnel and readiness management.
-
While many organizations prioritize analytics as a critical capability, it can be challenging to determine the current state and where to invest resources in the future. Since 2014, CANA Advisors has performed over twenty analytics capability evaluations using the INFORMS Analytics Maturity Model Survey. This presentation shares overall trends, comparisons, and lessons learned from the application of this tool at numerous government and commercial organizations. We also present a case study showing one government organization’s growth and development throughout five years of assessments. If you have ever asked yourself, “now that my organization now what?”, then this brief is for you.
Walter DeGrange
-
This presentation orients on the future of our Military OR profession. Where do we need to be going? What should we be doing? Strategic Analytics, the alignment of OR methods and management models with the ends-ways-means strategy paradigm is introduced. Foundational building blocks are presented: decision support capabilities, engineering systems, dynamic strategic planning, “engines for innovation”, and analytical architectures. Three recent applications of Strategic Analytics to Department of Defense enterprise system challenges are described.
Tina Diao
Tina Diao is a 3rd-year Ph.D. student studying decision and risk analysis in management science and engineering at Stanford University. Her current research interests lie in the preparedness and decision making under uncertainty in time-critical situations. She has work experiences in pension consulting and personal property & casualty insurance lines. Tina holds an M.S. in management science and engineering (technical and engineering management) from Stanford University and B.A.s in statistics and economics from University of California – Berkeley.
Richard Hun Kim
Richard Kim is a technical adviser to the United States Air Force. He helped to deploy the Space Based Space Surveillance satellite system and also led the development of an advanced command and control system for the next generation space protection infrastructure for U.S. space forces. Richard earned a B.S. in mathematics from UCLA, an M.S. in systems engineering from the Naval Postgraduate School, and a Ph.D. in management science and engineering from Stanford University.
-
Risk-based security (RBS) aligns security resources with risk. TSA Precheck is the poster child for RBS in the United States. This presentation discusses the value offered by TSA Precheck and how it can be further enhanced. We also discuss opportunities for RBS in reimagining airport security over the next two decades.
Sheldon H. Jacobson
Sheldon H. Jacobson is a Founder Professor in the Department of Computer Science at the University of Illinois at Urbana-Champaign. Dr. Jacobson has been working on the design and analysis of aviation security systems using operations research and artificial intelligence models since 1995. He has been recognized with numerous awards for his research, including a Guggenheim Fellowship and the INFORMS Impact Prize for his contributions to risk-based security. His aviation security research has been published in a wide spectrum of journals, including Naval Research Logistics, Transportation Science, and Journal of Transportation Security. He is an elected Fellow of the Institute of Industrial and Systems Engineers (IISE), the Institute for Operations Research and the Management Science (INFORMS), and the American Association for the Advancement of Science (AAAS).
-
Intelligence, surveillance, and Reconnaissance (ISR) mission are performed using unmanned air vehicles (UAVs) in military applications. Area of Operation is divided into macro-regions based on monitoring requirement which is quantified with information gain. Turning radius is taken into consideration for the dynamic behaviour of UAV. The main contribution of our work is obtaining an optimized path with curvature for UAV to maximize the information gain in a time-constrained mission. A heuristic approach is designed to improve quality and processing time of solutions for larger instances.
Rajan Batta
Dr. Rajan Batta is a SUNY Distinguished Professor in the Department of Industrial and Systems Engineering, University at Buffalo. He has a Bachelor of Technology degree in mechanical engineering from the Indian Institute of Technology, Delhi and a Doctor of Philosophy degree in operations research from the Massachusetts Institute of Technology.
Batta has made a major impact in the area of urban operations research, a branch of Operations Research (OR) concerned with logistical and planning applications impacting services such as emergency response, disaster logistics, and national security.
-
The U.S. Army maintains a ready-to-deploy unit at a large base in the southeastern United States. This force, known as the Immediate Response Force (IRF), must be able to deploy worldwide as quickly as 18 hours after notification. After collecting data during several exercises and using stochastic modeling, results on the time it takes to prepare for combat, load soldiers on aircraft, and ready equipment for transport were calculated. Using simulation, several recommended changes were investigated, resulting in an overall reduction in time required to deploy the IRF. Model results, solutions already implemented, and future research will be shared.
Samuel Herbert
Captain Samuel Herbert is an instructor in the Department of Systems Engineering at West Point. He currently teaches courses in the Fundamentals of Engineering Design & Systems Management, Supply Chain Engineering, and a Professional Engineering Seminar. Prior to joining the faculty at West Point, Sam served in various roles for the U.S. Army throughout Europe, Southwest Asia, and the United States. He received his B.S. from West Point and an M.S. from the Georgia Institute of Technology.
Christine Krueger
Major Christine Krueger is an Assistant Professor in the Department of Systems Engineering at West Point. She currently teaches courses in the Computer-Aided Systems Engineering, System Dynamics, and Statistics for Engineers. In addition, Christine advises teams and individuals researching a variety of topics to include Women, Peace and Security initiatives in Africa and a Lean Six Sigma project. Prior to joining the faculty at West Point, Christine served as a UH-60 Blackhawk pilot and deployed to Bagram, Afghanistan in support of Operation Enduring Freedom. She received his B.S. from West Point in Engineering Management ,a M.S. from Drexel University in Project Management and a M.S. from Northeastern University in Operations Research.
-
In this era of IoT/mobile apps, huge quantities of surveillance data about consumer activities offer opportunities for economic and societal value creation. However, they equally open up channels for consumer privacy breaches. In this talk we detail the need and modalities of consensual privacy trading – a pioneering idea we introduce (and mathematically validate) that allows economically efficiently trading consumer personal information with their consent and return benefits, rather than non-consensually ‘stealing’ this data, as many commercially motivated firms do. The major societal impact behind such an idea is the improvement in privacy welfare between consumers and data gatherers.
Ranjan Pal, PhD
Ranjan Pal is a faculty member in the Electrical and Computer Engineering division of EECS at University of Michigan Ann Arbor. His research interests lie in engineering robust multi-disciplinary solutions for effective cyber-security and information privacy management, using tools from decision and the applied mathematical sciences. Ranjan has a PhD in computer science from University of Southern California’s Viterbi School of Engineering, and has been a research fellow in the Computer Laboratory at the University of Cambridge.
Track 3: Resilience I: Security
Track Chairs: David Alderson & Daniel Eisenberg, Naval Postgraduate School
Over the last fifteen years, the concept of resilience has become popular for characterizing behavior in the presence of stress. The term has been applied to behavior at the level of a system, organization, or individual, and has become increasingly used across a variety of disciplines, including material science, ecology, psychology, emergency management, engineering, and national security. This track takes a modern look at the concept of resilience as it applies to the ability of a system to continue to operate in the presence of disruptive events and/or surprise.
-
Operations Research has yielded remarkable progress in the planning, operation, and design of a variety of civilian and military systems, making it possible to do things faster, better, and cheaper. However, the growing complexity of the world is showing that increased efficiency often comes at a price—when things fail, they can do so catastrophically. The term “resilience” has recently become a popular buzzword for systems that can absorb, resist, and recover from disruptive events. However, much of the work to date on this topic is merely descriptive and not particularly informative about what to do to make systems more resilient. This talk takes a modern look at the concept of resilience as it applies to the ability of a system to continue to operate in the presence of disruptive events and/or surprise. We also describe the limitations of big data analytics for resilience when systems are challenged by fundamental surprises never conceived during model development. In these cases, adoption of big data analytics may prove either useless for decision support or harmful by increasing dangers during unprecedented events.
David Alderson
-
Realized extreme weather events over the last decade and ever-increasing concern over potential cyber and physical incursions by our adversaries is driving changes in US government R&D portfolios focused on critical infrastructure resilience – and in particular critical energy infrastructure, i.e., power grid and natural gas pipeline networks. Specifically, emphasis is transitioning from resilience quantification (metrics) to advanced computational methods for situational awareness and operations planning. This transition is observed in key programs with national visibility, e.g., the US Department of Energy’s North American Energy Resilience Model (NAERM) initiative. From the viewpoint of optimization, this transition is conceptually natural – given a resilience metric, optimization enables proactive analytics to maximize resilience. However, this transition is very difficult to implement in the context of national-scale infrastructure resilience. This talk will highlight key R&D challenges associated with regional and national-scale computational resilience problems, emphasizing (1) key challenges for optimization, including scalability and multi-scale modeling; (2) the integral role of data science and its link to optimization; and (3) the critical interplay between simulation and optimization.
Jean-Paul Watson
Dr. Jean-Paul Watson is a Senior Research Scientist at Lawrence Livermore National Laboratory in Livermore, CA. He leads research projects in the area of optimization under uncertainty, with application to national security and critical infrastructure problems. He is co-founder of the widely used and R&D 100 award-winning Pyomo (www.pyomo.org) open source software package for mathematical optimization, a recipient of the INFORMS Computing Society Prize, and a INFORMS Franz Edelman Award finalist.
-
The rush to develop indicators and measures of infrastructure resilience before development of a more complete theory of resilience has resulted in an emphasis on things that can be counted, because those are amenable to measurement. The problem with this approach is that adaptive capacity resides not only in the resources that can be counted, but also in the capability to activate or deploy these resources. Thus, assessing resilience requires more than counting nouns. It requires description of verbs, including sensing, anticipating, adapting and learning. Because verbs can only be observed in action resilience engineering presents a serious challenge to existing paradigms of modeling and management. This presentation describes the relationship between language and resilience, and the challenge of understanding resilient in what complex, adaptive organizations do, in addition to what they have.
Thomas Seager
-
The advent of DevOps more than a decade ago has brought a revolution in the way that Critical Digital Services—from streaming services to online transactions—are developed and deployed. This creates new opportunities for service provisioning but creates new challenges to manage the reliability and risks of these services. In the past five years, software engineering and operations communities in the world of Internet-native applications have been discovering, exploring, and finding new connections between their domain and the field of Resilience Engineering. What has emerged is a unique synthesis of perspectives — on both theory and practice — that has great potential to influence the design and operations of not just the technical architectures of Critical Digital Services but the organizational efforts that support and extend them as the need to adapt arises. This presentation will provide some context around this emerging mix of domain expertise and research.
John Allspaw
-
Nicholas Judson
Dr. Nicholas Judson is the Assistant Leader of the Energy Systems Group at MIT Lincoln Laboratory. His group focuses on Department of Defense (DoD) tactical energy systems, DoD installation energy systems, domestic energy resilience, and advanced energy technology. He leads teams focusing on energy resilience analysis, planning, and exercises for the DoD to advance the capabilities of the DoD to pursue its mission during adverse events. He is a technical advisor for the ESTCP program within the DoD helping guide research and development projects to increase military energy resilience and energy security using advanced power generation options or control systems. He received his Ph.D. in Microbiology and Molecular Genetics from Harvard University and his A.B. in Biochemistry from Columbia University.
Track 3A: Risk in Cyberspace
-
Current organizational structures have proven insufficient for cyber and information assurance. The proposed supply chain audit and assessment framework will better support systems security and information compliance. The solution resides in technical workflows, databases, and machine learning to inform cybersecurity engineers and decision makers across the lifecycle of the equipment–from initial requirements specification, through pre-purchase audits, to deployment, maintenance, obsolescence, and destruction.
Randy Maule
Track 4: Resilience II: Infrastructure Focus
How can network systems – such as logistics, transportation, energy, communications, and even human networks – be better designed and protected to increase their robustness and resiliency to attack and degradation?
-
The U.S. is one of the most natural disaster-prone countries in the world. Since 1980 there has been 246 weather and climate disasters exceeding $1.6 trillion in remediation. In the last decade, the frequency of disaster events and their costs are on the rise. Complicating the impact of natural disasters is the population shift to cities and coastal areas which concentrates their effects. The need for governments and communities to prepare for, respond to, and recover from disasters is greater than ever before. But how can technologies like cloud, AI, and blockchain be leveraged across all aspects of the disaster management life cycle? This presentation will briefly address these questions and more.
Jeff Talley
Lt. Gen. (Ret) Jeff Talley is a dynamic academic, business, and government leader with over 38 years of global experience in large-scale organizational leadership, geopolitics, data/analytics and technology, and the environment. He currently serves as Vice President, Global Public Sector and Global Fellow, Center for the Business of Government, IBM. He is also a Professor and Scholar-in-Residence at the University of Southern California (USC). General Talley received his Ph.D. in Engineering from Carnegie Mellon University and an Executive M.B.A. from the University of Oxford. He is a registered Professional Engineer (P.E.), a Board-Certified Environmental Engineer (BCEE) in Sustainability, and a Diplomate, Water Resources Engineer (D.WRE).
-
Society depends decisively on the availability of infrastructures.
Even selective disruption may result in complex disruptions.
This talk gives an overview on the simulation and optimization of complex networks. Methods like Systems Engineering, Risk Management, Computational Intelligence, System Dynamics and Predictive Analytics are presented to master such complex adaptive networks via modern command & control systems.
We present results of the international projects RIKOV, REHSTRAIN and LIVE, and describe an outlook to the SANCTUM+ vision. At the end we describe an AI-based platform “IRIS” (Intelligent Reachback Information System IRIS) for combining “Digitization and Security”.Stefan Pickl
Prof. Dr. Stefan Pickl is Chair of Operations Research at University Bundeswehr München.
He is founding director of the Competence Center for Operations Research COMTESSA
for Safety and Security. He holds an Honorary Professorship at University of Nottingham,
Malaysia Campus. He is vice-chair of the German Committee for Desaster Reduction DKKV.
-
While AI is being applied in security, this paper presents further potential applications of AI in security. AI should be a good fit security problems as AI is ideally suited to analyze a) big data, b) complex decision-making spaces, and c) situations where there is ambiguity. We present high-level algorithms and methodologies for i] cost modeling, ii] evaluating effectiveness, and iii] estimating cost-effectiveness. Although applicable to security in general, these examples are from cybersecurity. The perspective is that of decision support, expert systems and knowledge bases rather than machine learning. We discuss future directions and provide a bibliography for further reading.
Soumya Moitra
Soumyo Moitra is a Senior Member of Technical staff at the Software Engineering Institute at Carnegie Mellon University. He has worked in many areas of cyber security including network traffic analysis and sensor deployment. His main interest is in developing metrics for cybersecurity management. He has published many articles on applications of OR and modeling in several fields.
-
Security professionals with constrained resources find it ever more challenging to select the best controls to detect, avoid, counter, and mitigate the effects of an expanding range of threats. We present an approach based on probabilistic risk and decision analysis that extends the widely-used FAIR framework. It lets you quantify the uncertainty in control costs and effectiveness against a range of threats. You will learn practical methods and software tools to assist in evaluating and selecting a cost-effective portfolio of controls.
Max Henrion, PhD
Max Henrion is CEO of Lumina Decision Systems and originator of the Analytica software. He has led decision analysis consultations private and public sector clients. He was formerly Professor at Carnegie Mellon and Consulting Professor at Stanford. He has coauthored three books and over 70 articles on risk, uncertainty, decision analysis, and artificial intelligence. He has a BA from Cambridge University. and PhD from Carnegie Mellon. He led a project that won the 2014 Decision Analysis Practice award. He was awarded the 2018 Frank Ramsey Medal from the INFORMS Decision Analysis Society.
-
This research presents a comprehensive data-driven approach to analyze the relationship among three groups of variables: (a) border security outcome metrics, input border security countermeasure metrics, and (c) external/environmental metrics. This project provides some novel insights on quantifying the effectiveness of border security counter-measures, and thereby facilitate more informed resource allocation in preventing drug/weapon smuggling, illegal migration, and terrorism.
Jun Zhuang
Dr. Jun Zhuang is a Professor and Director of the Decision, Risk & Data Laboratory, Department of Industrial and Systems Engineering at the University at Buffalo. Dr. Zhuang’s long-term research goal is to integrate operations research, big data analytics, game theory, and decision analysis to improve mitigation, preparedness, response, and recovery for natural and man-made disasters. Dr. Zhuang has been a principal investigator of over 30 research grants funded by the U.S. National Science Foundation (NSF), by the U.S. Department of Homeland Security (DHS), by the U.S. Department of Energy, by the U.S. Air Force Office of Scientific Research (AFOSR), and by the National Fire Protection Association.
-
In order to identify, among the myriad infrastructure assets in a region, the most critical ones to support a common objective, we view these assets as a network of interconnected nodes and quantify the operational dependencies between them. By applying the Functional Dependency Network Analysis (FDNA) methodology, this permits a quantitative assessment of significant impact on accomplishing the stated objective if these critical assets are compromised. Based on these findings, we also propose remedies to protect the critical assets and mitigate the adverse impact due to their loss.
Edward Hua, PhD
Dr. Edward Hua is a Simulation & Modeling Engineer at the MITRE Corporation. He currently works on several projects for the FAA and the Army. Dr. Hua received his Ph.D in Electrical & Computer Engineering from Cornell University, where his research focused on Mobile Ad Hoc Networks (MANETs).
Track 5: Resilience III: Cyberspace & Beyond
How can network systems – such as logistics, transportation, energy, communications, and even human networks – be better designed and protected to increase their robustness and resiliency to attack and degradation?
-
Securing cyber-physical systems is of paramount importance, but rigorous techniques to support decision-makers for high-consequence decisions are missing. The need for bringing rigor into cybersecurity is well-recognized, but little progress has been made over the last decades. We introduce our approach that starts with cyber experimentation to model the effect of a cyber attack on the power grid. Then we study how we can improve resilience of the physical system given the potential effects of a cyber attack, and prioritize cyber defenses based on the resilience of the physical system.
Ali Pinar
Ali Pinar is a distinguished member of technical staff at Sandia National Laboratories at Livermore, California. His work focuses on combinatorial problems in data mining and security applications. He currently leads Sandia’s efforts in rigorous cyber experimentation.
-
The concept of system resiliency in relation to cyber capabilities is important to individuals, businesses, and organizations alike. Currently, there are more qualitative metrics than quantitative metrics in measuring cyber resiliency. This paper offers a method of quantitatively measuring cyber resiliency using a multi-response optimization approach, in which the stakeholder in question may select from various measures to accommodate their preferences and goals.
Nerissa Siwietz
-
Risk assessment of critical infrastructures to cyber malfeasance needs to combine potential damage assessments to the infrastructure with cyber-security metrics derived from the digital framework controlling it. We discuss how customary path length metrics associated with attack trees are inadequate, and consider ways of assessing the cyber threat in ways and in units that directly support analysis of the risk of damage to the infrastructure.
David M. Nicol
David M. Nicol is the Franklin W. Woeltge Professor of Electrical and Computer Engineering at the University of Illinois at Urbana-Champaign. He serves as the Director of the Information Trust Institute, and Director of the Advanced Digital Sciences Centre (Singapore). He is Editor in Chief of IEEE Security and Privacy, was elected Fellow of the IEEE and of the ACM for contributions in parallel simulation, and is the inaugural recipient of the ACM SIGSIM Distinguished Contributions award.
-
Electric power system threats of a cyber-physical origin have the potential to significantly disrupt the operation of the electric grid. Power systems are cyber-physical critical infrastructure systems; the function of most other critical infrastructures furthermore typically assume reliable grid operation. We discuss the challenges and opportunities for enabling large scale cyber-physical power system modeling, followed by the use of these cyber-physical models to improve cyber-physical situational awareness and online intrusion tolerance and response in power systems.
Luis Garcia
-
Cybersecurity data science (CSDS) emerges from the practice of applying data science to prevent, detect, and remediate expanding and evolving cybersecurity threats. Based upon interviews with 50 global cybersecurity data scientists, this presentation offers guidance to practitioners and managers interested to advance CSDS professional practice and general effectiveness. This includes those planning operational implementations and/or organizational programs.
Scott Mongeau, PhD, CAP
Scott Mongeau (INFORMS CAP, MA, MA, GD, MBA, PhD (abd)) is a Cybersecurity Data Scientist – Principal Business Solutions Manager at SAS Institute. He has more than three decades of experience designing and deploying data analytics solutions for a range of industries, including management consulting (Deloitte and own firm SARK7), intelligence services / law enforcement / military, telecom, financial services, IT, bio-pharma, and start-ups. Scott is in the process of defending his PhD dissertation ‘Cybersecurity Data Science: Best Practices in an Emerging Profession’ at Nyenrode Business University in the Netherlands, which will result in a forthcoming book published by Springer.
-
In this presentation, we will discuss how organizations make cybersecurity decisions, how they currently prioritize their cybersecurity risks, and how they decide what capabilities to invest in and apply limited resources to. Leveraging the Open SIPmath™ Standard from 501(c)(3) ProbabilityManagement.org, we will demonstrate a Cybersecurity Risk Management modeling framework to help Decision-Makers visualize resource investment decisions through analytically-based and interactive risk assessment models that allow Decision-Makers to rapidly visualize aspects of their risk and know which levers to pull to influence it.
Shaun Doheney, PMP, CAP
Shaun Doheney is the Chief Analytics Officer for JDSAT – a certified Service-Disabled Veteran-Owned Small Business specializing in Operations Research and Data Science. He is also the Chair of Resources and Readiness Applications at ProbabilityManagement.org – a nonprofit devoted to making uncertainty actionable. He is an INFORMS Certified Analytics Professional (CAP®) and PMI-certified Project Management Professional (PMP®) with a B.S. in Mathematics, an M.S. in Operations Analysis, and a Graduate Certificate in Data Analytics. As a Marine Corps Lieutenant Colonel (Retired) and Marine Operations Research Analyst, he has performed quantitative and qualitative analyses and evaluations across major DoD decision support processes.
-
Cybersecurity data science (CSDS) emerges from the practice of applying data science to prevent, detect, and remediate expanding and evolving cybersecurity threats. Based upon interviews with 50 global cybersecurity data scientists, this presentation offers guidance to practitioners and managers interested to advance CSDS professional practice and general effectiveness. This includes those planning operational implementations and/or organizational programs.
Theodore T. Allen
Theodore (Ted) Allen is an Associate Professor in the Integrated Systems Engineering department at the Ohio State University. He received his B.A. from Princeton, his M.S. from UCLA, and his Ph.D. from the University of Michigan (1997). He is currently the president of the Social Media Analytics section of INFORMS and the simulation area editor of Computers & Industrial Engineering (IF: 3.2). His papers on Gaussian Stochastic Process (GSP) and multi-fidelity optimization with manufacturing process design applications are highly cited. Overall, Dr. Allen has published over 60 refereed publications and received over 30 grants as PI including from NSF, EWI, ARCYBER, Ford, and GE Appliances. Dr. Allen’s textbook on engineering statistics is in its third edition. His new textbook (under preparation) will explain how machine learning including reinforcement learning can aid in process parameter design, manufacturing 5.0 decision support, and related cyber security decision-making. He has also served as associate editor for the Journal of Manufacturing Systems and Quality Approaches in Education and as a reviewer for Operations Research, EJOR, and Technometrics.
Track 6: Disaster Preparedness
How can O.R. and analytics better help government agencies and the public writ large understand and prepare for emergencies?
-
Evaluating Disaster resilience capacity are very import for disaster- sensitive areas in the world. An effective disaster management plan will reduce life and property losses. In this study, the earthquake resilience capacity of any city and her districts is evaluated by data envelopment analysis (DEA) and some dea related models. Istanbul is chosen as a study area. She one of the most important centers of Turkey and the world. On the other hand, Istanbul is also an disaster-prone city, and has experienced many of them throughout history. The efficient and inefficient areas of studied areas are determined and discussed, for possible improvements of inefficient units in input and output values.
Abdullah Korkut Ustun
-
In September 2018, Hurricane Florence hit the Carolinas. More than a million people were forced to evacuate as rainfall hit and wind speeds amplified. In response, the National Guard immediately started to collect imagery from airborne platforms and used it to identify flooded areas and damaged infrastructure. JHU APL was tasked by the Joint AI Center to accelerate and automate this process using machine learning. In this talk, we will provide details about how we developed a scalable software capability to process large amounts of imagery and the lessons learned from our iterative design discussions with the National Guard Unclassified Processing, Assessment and Dissemination (UPAD) analysts.
Beatrice Garcia
-
We undertake the most comprehensive survey to-date of the emergency response operations research literature and identify 644 papers.
- Using unsupervised learning techniques, we identify a number of distinct research clusters that have developed in isolation from one another.
- Based on axioms revealed in our study we propose a modelling framework that captures the essential elements of emergency response decision making.
James P. Minas, PhD
Dr. James P. Minas is an Assistant Professor of Business Analytics and Decision Sciences in the School of Business at Ithaca College. His research interests include emergency preparedness and response and supply chain resilience, with publications appearing in European Journal of Operational Research, Annals of Operations Research, Decision Support Systems and the International Journal of Wildland Fire. Dr. Minas received a PhD from the Royal Melbourne Institute of Technology (RMIT), completing his doctoral research in conjunction with the Bushfire Cooperative Research Centre.
-
Some emergency supply planning challenges, like dynamic post-disaster environments and traffic congestion, are seldom addressed in the literature. We propose a stochastic programming model, which integrates pre-disaster supplies pre-positioning and transportation network development decisions with post-disaster transportation decisions. We solve the formulated mixed-integer nonlinear program and its variant using generalized Benders decomposition algorithms. Moreover, some case study results are presented.and/or organizational programs.
Xiaofeng Nie, PhD
Xiaofeng Nie received a Ph.D. degree in operations research from the University at Buffalo, Buffalo, NY, USA, in 2008. He is an Associate Professor with the Department of Engineering Technology and Industrial Distribution, Texas A&M University, College Station, TX, USA. His research interests include humanitarian logistics, homeland security, and supply chain risk management.
-
The U.S. Department of Homeland Security (DHS), with support from Booz Allen Hamilton, fosters data-driven decision-making to inform policy implementation, improve operational practices, and drive data science capabilities. We will share best practices for integrating data science into government culture, data science and analytics for resource allocation and strategic decision-making, and recommendations for the future of government data analytics.
Tim Caudill
Tim Caudill is the Strategy and Operations Analysis (SOA) Chief leading innovative analytics, data science, research, and modeling or forecasting capabilities for the U.S. Department of Homeland Security (DHS). Mr. Caudill has over 21 years of federal service leading technical programs for DHS and the Department of Defense (DoD). He is a U.S. Navy veteran and is pursuing a PhD in business administration focused on engineering/technology management.
-
Service liability interconnections among modern IT/IoT-networked service organizations create potential channels for (catastrophic) cascading service disruptions due to cyber-risks posed by state-of-the-art DDoS, APT, and ransomware hacks. In this talk we discuss the role of the liability network topology and statistical cyber-risk distributions on the feasibility and effectiveness of the growing and much needed risk management services (e.g., cyber-insurance) for modern cyber/critical-infrastructures, and provide cautionary suggestions for the development of such markets for the welfare of society.
Ranjan Pal, PhD
Ranjan Pal is a faculty member in the Electrical and Computer Engineering division of EECS at University of Michigan Ann Arbor. His research interests lie in engineering robust multi-disciplinary solutions for effective cyber-security and information privacy management, using tools from decision and the applied mathematical sciences. Ranjan has a PhD in computer science from University of Southern California’s Viterbi School of Engineering, and has been a research fellow in the Computer Laboratory at the University of Cambridge.
-
Katrina and Fukushima show that the developed world faces disaster relocation, with high economic/social costs. Economic costs are nonlinear in both magnitude and duration of relocation. A brief evacuation may be minimally disruptive; a few months would be much more disruptive, but eventually, costs would diminish. Costs can be expected to increase more than linearly in the number of people relocated, and depend on the nature of interdicted assets. Disasters can result in the need to relocate a million people or more, arguing for research to improve preparedness, resilience, and recovery.
Vicki M. Bier
Vicki Bier holds a joint appointment in the Departments of Industrial and Systems Engineering and Engineering Physics at the University of Wisconsin-Madison. She has directed the Center for Human Performance and Risk Analysis (formerly the Center for Human Performance in Complex Systems) since 1995, and chaired the Department of Industrial and Systems Engineering from 2011 to 2016. She received a B.S. in Mathematical Sciences from Stanford University in 1976, and a Ph.D. in Operations Research from the Massachusetts Institute of Technology in 1983.
Track 7: Frontiers I: Machine Learning & Modern Data Analytics
How can the security community benefit from and drive innovation and advances in operations research tools, techniques, and practices?
-
Sensors that record sequences of measurements are now embedded in many products. There is information in the shapes of the sensor stream that is highly predictive of the likelihood of a product failure or batch yield. These data are often being used inefficiently due to lack of knowledge and tools for how to properly leverage it. In this presentation we will show how to fit splines to data streams and extract features called functional principal component scores. Then, we use these features as inputs into machine learning models like neural networks. Answering a wide variety of questions becomes a two-step process of functional feature extraction followed by modeling using those features as inputs.
Thomas A. Donnelly, PhD, CAP
Tom Donnelly, PhD, CAP, is a Principal Systems Engineer for the SAS Institute where he supports users of JMP software for data visualization, exploration, and discovery in the Defense community. Over the past 35 years Tom has taught more than 300 industrial short courses to engineers and scientists in the field of Design of Experiments (DOE). Over the last 10 years he has taught more than 100 short tutorials on topics that also include; data visualization, predictive modeling, and machine learning.
-
By sequentially running High-Performance Computing (HPC) simulations in blocks of space-filling DOE trials, one can efficiently increase the accuracy of fast surrogate models. Factor ranges can be shrunk between blocks to focus new trials in regions of complex behavior. When a sufficiently accurate surrogate model has been achieved – using machine learning methods – HPC resources can then be released to other M&S projects. Surrogate models can be used to do sensitivity and trade-space analysis of the factors. The surrogate models can provide accurate predictions – instantaneously – for never-before proposed scenarios that fall within the ranges of the long-running simulation’s factors.
Thomas A. Donnelly, PhD, CAP
Tom Donnelly, PhD, CAP, is a Principal Systems Engineer for the SAS Institute where he supports users of JMP software for data visualization, exploration, and discovery in the Defense community. Over the past 35 years Tom has taught more than 300 industrial short courses to engineers and scientists in the field of Design of Experiments (DOE). Over the last 10 years he has taught more than 100 short tutorials on topics that also include; data visualization, predictive modeling, and machine learning.
-
Machine Learning Operations (MLOps) is an emerging area of importance in the application of data science methods to real-world problems. This presentation provides a working definition of the term, lays out guidelines for its implementation, and demonstrates one set of tools to enable effective MLOps. In particular the presentation will describe how Cloudframe applies these principles in the Defense, Intelligence, and Homeland Security domains using an open source technology stack.
Robert Lantz, CAP
Rob Lantz is the CEO and co-founder of Cloudframe Analytics, a data science firm that builds Machine Learning backed web applications for public and private sector clients. He holds a Master of Science in Operations Research and is a 10 year veteran of the US Marine Corps. He enjoys spending time with his family, the outdoors, rooting for the Washington Nationals, and all combinations thereof.
-
The ability to economically sample a group and make inferences on its correlation to the population is imperative for decision makers. Survey analyses are the means to convert this raw group data into useable information. Decision science tools and machine learning techniques are essential in bridging this information gap. But they come with an inherent risk to decision makers. Being misinformed about the operating environment from faulty analysis poses a true danger to the decision maker; it assumes confidence when there should be none. Clustering is widely effective in framing target sets. Understanding its utility and dangers is revealed in its application on national survey data-set.
Walter Kulzy
Walter W. Kulzy is a Senior Operations Research Analyst with Johns Hopkins University Applied Physics Lab. He previously served 21 years in the United States Navy as an Operations Research Analyst and Logistics expert. His experienced included Readiness Officer with SOCEUR, M&S team lead with USCENTCOM, Operations Planner with USSOCOM Wargame Center, Chief Analytics Officer with Naval Weapon Systems Support, and Logistics Support Lead with NSW DevGroup.
-
Mobile agents are generally applicable in military fields such as detection of nuclear, biological, and chemical (NBC), security, and targeting system. We develop a mobile multi-agent sensing problem mathematically to monitor a set of nodes and detect event occurrences from the nodes. This problem can be represented as a submodular maximization problem under a partition matroid constraint, which is NP-hard in general. Therefore, the optimal solution can be intractable to compute within a reasonable time. We propose decent algorithms based on the greedy algorithm and prove new efficiency ratios of the algorithms. Finally, we demonstrate the performance of our results through numerical tests.
Jongmin Lee
Jongmin Lee is a Graduate Student of Industrial Engineering at Seoul National University in Korea. His research interests are computer aided decision support systems, optimization, and supply chain management.
-
This talk will introduce new developments in statistical estimation on networks. We will focus on Fused Density Estimation, a technique for constructing density estimates from observations which occur along the edges of a network. We will address computational challenges and provide theoretical support for the use of the Fused Density Estimation. Lastly, we review applications of the technique to infrastructure networks, demonstrating the efficacy of the technique with a number of real-world examples.
Robert L. Bassett
Dr. Robert L Bassett is an Assistant Professor of Operations Research at the Naval Postgraduate School. His research interests include both theoretical and computational aspects of data-driven decision making, with emphasis on defense applications. Before joining NPS, Dr. Bassett worked with Sandia National Laboratories and as a research mathematician for the US Department of Defense.
Track 8: Frontiers II: Interdisciplinary Applications
-
Game theory is an essential analytical tool for numerous security applications. In this presentation, we demonstrate, through a series of novel network models, how game theory can also assist in disaster preparedness and response of organizations. We describe both deterministic and stochastic models that integrate financial and logistical elements and that consist of multiple organizations, freight service providers, and points of demand. The models capture the behavior of decision-makers and yield solutions that help victims while reducing materiel convergence. Case studies on real-life disasters demonstrate the efficacy of the framework and provide managerial insights.
Anna B. Nagurney
Dr. Anna Nagurney is the John F. Smith Memorial Professor of Operations Management at the Isenberg School of Management at UMass Amherst and the Director of the Virtual Center for Supernetworks. She is a Fellow of INFORMS, of RSAI, and of the Network Science Society. Anna is the author/co-author of over 200 journal articles, 50 book chapters, and 14 books. She holds a PhD from Brown University. Anna has been a Science Fellow at the Radcliffe Institute for Advanced Study at Harvard University, a Visiting Fellow at All Souls College at Oxford University, and a Fulbrighter twice (in Austria and Italy), among other appointments. Prior to her career in academia, she worked as a software/technical consultant for two companies supporting the Naval Underwater Systems Center in Newport, Rhode Island. Her first ever conference presentation was at NPS in Monterey.
-
Today’s cyber threats to our families, communities, and nation are numerous. This talk will cover research conducted between Arizona State University and the U.S. Army over the last three years – imagining the cyber threats that will emerge over the next decade, determining new threat actors that will emerge, and starting to work with industry, academia, and government to proactively change. Based on science fact but illustrated as graphic novellas, we will paint a vision of the future that will inspire today’s generation to act based on the Threatcasting analytical methodology.
Natalie Vanatta
Natalie Vanatta is a U.S. Army Cyber officer and an Academy professor at the Army Cyber Institute. Here she focuses on bringing private industry, academia, and government agencies together to explore and solve cyber challenges facing the U.S. Army in the next 3-10 years. She holds a PhD in applied mathematics as well as degrees in computer engineering and systems engineering. Natalie has also served as a Distinguished Visiting Professor at the National Security Agency and the technical director to Joint Task Force Ares. Currently, she is serving as the team leader for 01 National Cyber Protection Team as part of the Cyber National Mission Force.
-
I will discuss ways to assess important counter-terrorism values by eliciting and aggregating the judgments of experts. Defensive budget allocations depend on the attacker’s valuations of various attack targets and scenarios, but those valuations are uncertain and multi-attribute. Attackers may care about property damage, fatalities, infrastructure, publicity, etc. Also, different attackers may have different goals (maximizing vs. minimizing fatalities), and preferences may be non-monotonic (e.g., enough fatalities to gain publicity, but not enough to inspire retaliation). Assessing utility functions for terrorists is difficult, but mathematical techniques can help in this process.
Vicki M. Bier
Vicki Bier holds a joint appointment in the Departments of Industrial and Systems Engineering and Engineering Physics at the University of Wisconsin-Madison. She has directed the Center for Human Performance and Risk Analysis (formerly the Center for Human Performance in Complex Systems) since 1995, and chaired the Department of Industrial and Systems Engineering from 2011 to 2016. She received a B.S. in Mathematical Sciences from Stanford University in 1976, and a Ph.D. in Operations Research from the Massachusetts Institute of Technology in 1983.
-
This talk will present a warning systems model in which early-stage cyber threat signals are generated using machine learning and artificial intelligence (AI) techniques. Cybersecurity is most often, in practice, reactive. Based on the manual forensics of machine-generated data by humans, security efforts only begin after a loss has taken place. The current security paradigm can be significantly improved. Cyber-threat behaviors can be modeled as a set of discrete, observable steps called a ‘kill chain.’ Data produced from observing early kill chain steps can support the automation of manual defensive responses before an attack causes losses. However, early AI-based approaches to cybersecurity have been sensitive to exploitation and overly burdensome false positive rates resulting in low adoption and low trust from human experts. To address the problem, this research presents a collaborative decision paradigm with machines making low-impact/high-confidence decisions based on human risk preferences and uncertainty thresholds. Human experts only evaluate signals generated by the AI when decisions exceed these thresholds. This approach unifies core concepts from the disciplines of decision analysis and machine learning by creating a super-agent. An early warning system using these techniques has the potential to avoid more severe downstream consequences by disrupting threats at the beginning of the kill chain. Real world data collected from a network of honeypots is used to demonstrate this concept.
Isaac Faber
Isaac is currently serving as the Chief Data Scientist of the U.S. Army’s Artificial Intelligence Task Force where he oversees the technical aspects of the projects in the portfolio. He was also the lead data scientist at U.S. Army Cyber Command and led the architecting and deployment of the first operational big data platform in the DoD. He has a Ph.D. from Stanford where he researched risk management in cybersecurity using AI and human cooperation.
-
The metalog probability distributions can represent virtually any continuous shape with a single family of equations, making them far more flexible for representing data than the Pearson and other distributions. The metalogs are easy to parameterize with data without non-linear parameter estimation, have simple closed-form equations, and offer a choice of boundedness. Their closed-form quantile functions (F-1) enable fast and convenient simulation. In security and other applications, the metalog distributions are more flexible, widely applicable, and easier to use than traditional probability distributions.
Thomas W. Keelin
Tom Keelin is Managing Partner of strategy consulting firm Keelin Reeds Partners, LLC, and general partner of real estate fund Millennial Capital, LLC — both of which motivated his invention and reduction to practice of the metalog probability distributions. Tom is fellow of Society of Decision Professionals, a founder and director of the Decision Education Foundation, and previously Worldwide Managing Director of the Strategic Decisions Group. He holds three degrees from Stanford University: BA in economics, and MS and PhD in engineering-economic systems.
-
This presentation orients on the future of our Military OR profession. Where do we need to be going? What should we be doing? Strategic Analytics, the alignment of OR methods and management models with the ends-ways-means strategy paradigm is introduced. Foundational building blocks are presented: decision support capabilities, engineering systems, dynamic strategic planning, “engines for innovation”, and analytical architectures. Three recent applications of Strategic Analytics to Department of Defense enterprise system challenges are described.
Colonel Greg H Parlier, PhD, PE
Dr. Greg H. Parlier is a West Point graduate, combat veteran, and retired US Army Colonel with assignments including 8 years in the 82nd Airborne Division, five operational deployments, and service in over 20 foreign countries. With over 40 years experience in leadership and management, operations research (OR) and organizational innovation, strategic planning and defense analysis, his unique qualifications and expertise are building, developing, and leading successively larger multi-disciplinary teams confronting increasingly more demanding transformational challenges in large commands and complex organizations. His contributions include teaching as an academy professor, research and publications as a university research scientist, and professional society leadership at local, state, regional, national, and international levels.
-
Increasingly, social media has been used as a platform for providing timely crisis communication during a disaster. Unfortunately, rumor spreading, and in particular, rumors that are later proven to be untrue, have been identified as critical issues for the use of social media during disasters. Millions of tweets posted during Hurricanes Harvey, Irma, and Sandy are collected and analyzed. We also investigate several cases of rumor responding and debunking behaviors of Twitter users. This research provides some novel insights on crisis communication and rumor management using social media during disasters.
Jun Zhuang
Dr. Jun Zhuang is a Professor and Director of the Decision, Risk & Data Laboratory, Department of Industrial and Systems Engineering at the University at Buffalo. Dr. Zhuang’s long-term research goal is to integrate operations research, big data analytics, game theory, and decision analysis to improve mitigation, preparedness, response, and recovery for natural and man-made disasters. Dr. Zhuang has been a principal investigator of over 30 research grants funded by the U.S. National Science Foundation (NSF), by the U.S. Department of Homeland Security (DHS), by the U.S. Department of Energy, by the U.S. Air Force Office of Scientific Research (AFOSR), and by the National Fire Protection Association.
