DIMACS/DyDAn/LPS Workshop on Port Security/Safety, Inspection, Risk Analysis and Modeling

November 17 - 18, 2008
DIMACS/DyDAn Center, CoRE Building, Rutgers University

Organizers:
Tayfur Altiok, Rutgers University, altiok at rci.rutgers.edu
Tami Carpenter, DIMACS/DyDAn, tcar at dimacs.rutgers.edu
Benjamin Melamed, Rutgers University, melamed at rbs.rutgers.edu
Fred Roberts, DIMACS/DyDAn, froberts at dimacs.rutgers.edu
Presented under the auspices of the DIMACS Center, the Center for Dynamic Data Analysis (DyDAn), the Laboratory for Port Security (LPS) and the DIMACS/Los Alamos National Laboratory Partnership on Algorithms for Port-of-Entry Inspection.

Abstracts:


Tayfur Altiok, ISE, Rutgers University

Title: CAIT/DIMACS Laboratory for Port Security

The Laboratory for Port Security (LPS) is a is a joint effort of DIMACS and the Center for Advanced Infrastructure and Transportation (CAIT), a U.S. Department of Transportation-designated University Transportation Center. LPS collaborates and coordinates activities with key local, state, and federal agencies. LPS's mission is to improve port security by researching the complexities of cargo vessel traffic and container inspection processes, developing workable solutions, and supporting their implementation. LPS conducts research, education, and training projects that concentrate on the logistics involved in:


Tayfur Altiok, ISE, Rutgers University

(joint work with Birnur Ízbas, Bogaziši University)

Title: Risk Analysis of the Transit Vessel Traffic in the Strait of Istanbul

In this study, we have analyzed safety risks pertaining to transit vessel maritime traffic in the Strait of Istanbul and proposed ways to mitigate them without causing additional vessel delays. Actual vessel traffic operations and data were studied and a simulation model was developed to mimic the operation with its geographical and weather dynamics. Safety risk analysis was performed by incorporating a probabilistic accident risk model into the simulation model. A scenario analysis was carried out to study the behavior of the accident risks, with respect to changes in the surrounding geographical, meteorological and traffic conditions. This analysis allowed us to investigate the impact of various factors on the risk profile of the Strait. These factors included vessel arrival rates, scheduling pilotage and overtaking policies, and local traffic conditions.


George Apostolakis, Nuclear Science and Engineering, Massachusetts Institute of Technology

(joint work with Michael Shattan, Massachusetts Institute of Technology)

Title: An Analytic-Deliberative Process for the Selection of Radiation Detection Systems for Shipping Ports and Border Crossings

We propose an Analytic-Deliberative Process (ADP) for the selection of radiation portal monitors that allow radiological screening of containers entering the country without detrimental effects on the stream of commerce through the terminals. We conduct a case study with four stakeholders using the ADP. The process results in a preference ranking of the decision options according to their value to each stakeholder. The analytical results are then used to structure a deliberation in which the four stakeholders use both the analytical results and any pertinent information outside the analysis to create the final ranking of the decision options. The ADP format captures and quantifies explicitly subjective judgments affecting the final decision. This process facilitates discussion and enhances consensus building.


Vicki Bier, University of Wisconsin-Madison

Title: How Many Containers to Inspect to Deter Terrorist Attacks

I will present a model for identifying the optimal number of containers to inspect in order to minimize the defender's expected loss, using game theory. The model considers multiple types of attackers, each deciding whether to smuggle a particular type of weapon into the U.S. Results show that with a sufficiently high detection probability (and a sufficiently high smuggling cost to the attacker), attackers can be deterred with less than one hundred percent inspection. potential impact, focusing specifically on the use of supply chain inventory as a risk mitigation strategy for a one supplier, one customer system in which goods are transported through a port subject to temporary closures. Closure likelihood and duration are modeled using a completely observed, two-state, exogenous Markov chain. Order lead times are dependent on the status of the port, including potential congestion backlogs of unprocessed work. We develop an infinite-horizon, periodic-review inventory control model for determining optimal average cost ordering policies under linear ordering costs with backlogged demand, and a method for evaluating policy cost. When congestion is negligible, the optimal policy is state invariant. In the more realistic case of non-negligible congestion, this result no longer holds. For the scenarios considered, numerical results indicate that operating margins may decrease 10% for reasonable-length port closures, margins may be eliminated completely without contingency plans, and expected holding and penalty costs may increase 20% for anticipated increases in port utilization.


Cherrie Black, Chief, Critical Infrastructure Bureau, NJ Office of HS and Preparedness

Title: Critical Infrastructure Modeling and Simulation: A State Perspective on Current Needs and Challenges

The talk will encompass a description of NJ's current infrastructure protection and analysis priority: Interdependency analysis, modeling, simulation and resiliency planning for the Port of New York and New Jersey and the seven county Northern New Jersey UASI region.


Endre Boros, RUTCOR and Paul Kantor, SCILS, Rutgers University

Title: Optimal sensor sequencing for container inspection

We consider the problem of combining a given set of diagnostic tests into an inspection system that can classify items with maximum accuracy subject to budget constraints. One motivating application is sequencing diagnostic tests for container inspection problems, where the diagnostic tests may correspond to radiation sensors, documents checks, or imaging systems. We propose a dynamic programming algorithm that enumerates all of the efficient, undominated, inspection policies in the cost-detection domain. Our inspection policies may sequence arbitrary number of tests and are not restricted in the branching factor (or number of channels). Our approach directly solves the bi-criterion optimization problem of maximizing detection and minimizing cost, and thus supports sensitivity analysis over different budget or detection requirements.


George Cummings, Director of Homeland Security, Port of Los Angeles

Title: Overview of the Security Program and Port Recovery Planning at the Port of Los Angeles

This talk will cover operational measures, interagency coordination, public/private coordination, use of technology, and funding issues for security and recovery planning at the Port of Los Angeles.


E. A. Elsayed, Industrial and Systems Engineering, Rutgers University

Title: Optimization of Containers Inspection at Port-of-Entry

Inspection of containers arriving at the port-of-entry is a challenging problem as both the number of containers and inspection attributes of the containers increase. The sequence of inspections and the level of inspection have a major impact on the total cost of inspection and delay of containers at a port. This talk presents methods that search for the optimum threshold levels at inspection stations as well as the optimum inspection sequence to minimize the total cost and delays.


Jeff Fuller, Director, HS Services, Teledyne Brown Eng., and Brady Downs, LCDR, USCG

Title: USCG Maritime Security Risk Management: Six Risk Management Cases

In 2006, the USCG rolled out MSRAM (Maritime Security Risk Analysis Model) as a program to provide Commanders at all levels from Captains of the Port and Sector Commanders to the Commandant with risk information to support tactical, operational and strategic decisions. The MSRAM program is a method, a process and a tool kit. The front line of MSRAM is the Port Security Specialist assigned to field units who attend annual training and assist their staffs and local Area Maritime Security Committees (AMSC) with the analysis of consequences and vulnerabilities of twenty-four different possible terrorist attack modes. The Intelligence Coordination Center provides the threat component of the risk formula.

To provide insight into the scope and impact of this program we summarize several risk stories that illustrate how the USCG has aggressively moved to reduce risk of terrorist attack. To protect security, the locations of these stories are not included.


Noah Gans, Wharton School, University of Pennsylvania

(joint with Nitin Bakshi, London Business School and Stephen Flynn, Council on Foreign Relations)

Title: Measuring the Operational Impact of Container Inspections at International Ports

A recent US law mandating non-intrusive imaging and radiation detection for 100% of US-bound containers at international ports has provoked widespread concern that the resulting congestion would hinder trade significantly. Using detailed data on container movements, gathered from two large international terminals, we simulate the impact of various inspection policies being considered. We find that the current inspection regime can only handle a small percentage of the total load. An alternate inspection protocol which emphasizes screening -- a rapid primary scan of all containers, followed by a more careful secondary scan of only a few containers that fail the primary test -- holds promise as a feasible solution to the 100% scanning requirement.


Gary Gaukler, Industrial and Systems Engineering, Texas A&M University

Title: A Modeling Framework for Detecting Illicit Nuclear Materials in Seaborne Containers

The goal of our research effort is to establish a system to prevent terrorists from smuggling nuclear materials into the United States. In this talk, we consider container traffic at a single sea port. We determine appropriate inspection policies to decide the level of scrutiny to use for any given shipment. To do this, we create a mathematical framework for evaluating and optimizing different inspection policies according to performance measures of interest, such as detection probability and system throughput. A novel feature of this framework is that it allows us to calculate the value of having prior information about the contents of containers (e.g from the Bill of Lading or through X-ray inspections).


Gariann Gelston, Pacific Northwest National Laboratory

Title: The Rail Test Center, Enhancing Intermodal Rail Security

The Rail Test Center (RTC), sited at the Port of Tacoma (POT), provides the Domestic Nuclear Detection Office (DNDO) with the ability to test radiological and nuclear threat detection systems associated with the unique environment of intermodal rail scanning. In this operational seaport setting, several aspects of risk naturally co-exist and need to be taken into account simultaneously when evaluating potential radiation detection solutions. Three critical areas of risk accounted for in the activities of the RTC are: 1) national security risk, 2) operational impact or business risks driven by the profit motive, and 3) technological risk associated with the radiation detection systems. The RTC project is addressing these three areas of risk and is using engineering principals, innovation, and key stakeholder relationships to define resolution pathways for mandatory radiation detection scanning at U.S. Seaports. This discussion will highlight where these risks areas interact, and the use of data and simulation modeling techniques to better understand and communicate impacts.


Michael Greenberg, Bloustein School of Planning and Public Policy, Rutgers University

Title: Risk Analysis: Theory & Practice

Risk analysis, which includes risk assessment and management, frequently is blamed for bad public policy decisions and the probably does not receive enough credit for good policy decisions. But no one denies the impact of this multi-disciplinary policy process on deliberative thinking about risk. I briefly review the elements of risk analysis, compare it to alternatives, and then use two examples to illustrate that is always can be helpful, but is never a panacea.


Nick Hengartner, Los Alamos National Laboratory

Title: Enabling Port Security using Passive Muon Tomography

This talk discusses how muons generated from cosmic rays can be used to detected and image shielded special nuclear material (SNM) in shipping containers. The detector measures the Coulomb scattering induced by high-Z material in a cargo. As a result, shielding the SNM increases the probability of detection.


Richard Hoshino, Canada Border Services Agency

Title: Applying cost curves to marine container inspection

Drummond and Holte introduced the theory of cost curves, a graphical technique for visualizing the performance of binary classifiers over the full range of possible class distributions and misclassification costs. In this talk, we apply cost curves to the risk-assessment of marine cargo containers. We illustrate how implementing even a basic predictive model would save our Agency both time and money, regardless of class distributions or misclassification costs.


Dorit Hochbaum, School of Business, University of California at Berkeley

Title: The multi-sensor nuclear threat detection problem

One way of reducing false-positive and false-negative errors in an alerting system, is by considering inputs from multiple sources. We address here the problem of detecting nuclear threats by using multiple detectors mounted on moving vehicles in an urban area. The likelihood of false alerts diminishes when reports from several independent sources are available. However, the detectors are in different positions and therefore the significance of their reporting varies with the distance from the unknown source position. An example scenario is that of multiple taxi cabs each carrying a detector. The real-time detectors' positions are known in real time as these are continuously reported from GPS data. The level of detected risk is then reported from each detector at each position. The problem is to delineate the presence of a potentially dangerous source and its approximate location by identifying a small area that has higher than threshold concentration of reported risk. This problem of using spatially varying detector networks to identify and locate risks is modeled and formulated here. The problem is then shown to be solvable in polynomial time and with a combinatorial network flow algorithm.


David Madigan, Statistics, Columbia University and Rutgers University

(joint work with Sushil Mittal and Fred Roberts, Rutgers University)

Title: Optimal Inspection Sequences

As a stream of containers arrives at a port, a decision maker must decide which "inspections" to perform on each container. Current inspections include neutron/gamma emissions, radiograph images, induced fission tests, and checks of the ship's manifest. The specific sequence of inspection results will ultimately result in a decision to let the container pass through the port, or a decision to subject the container to a complete unpacking. We formulate this sequencing task as a problem of finding an optimal binary decision tree for an appropriate Boolean decision function. We report on new algorithms that offer specific computational advantages. The improvements enable us to analyze substantially larger applications than was previously possible.


Isaac Maya, CREATE HS Center, University of Southern California

(joint work with Onur Bakir, Petros Ioannou, Michael Orosz, all of USC)

Title: Port Operations Modeling for Risk Management and Resource Allocation

This project is developing a simulation model of port operations for performing risk-based analysis of security countermeasures to reconcile the seemingly opposing goals of minimizing the risk of terrorism while maintaining unimpeded flow of daily port activity.

The Ports of Los Angeles and Long Beach (POLA/LB) are facing the challenge of conducting daily port activities with maximum focus on homeland security efforts. Inspection of incoming cargo and protection of port perimeters and waterways have become top priorities of port officials. There is a growing need to optimize allocation of resources on security investments and use technology effectively within the port complex to maximize the benefits of each dollar spent on homeland security missions.

We are developing a simulation modeling approach that incorporates detailed understanding of daily port operations to quantify the impacts of technological countermeasures on both risk and daily business operations. The novelty in our approach is in exploring optimal resource allocation strategies within a simulation model that will measure relative impacts of security activities and countermeasure on risk, business continuity and the pace of daily port activities.


Laura McLay, Statistical Sciences and Operations Research, Virginia Commonwealth University

Title: Screening Cargo Containers for Nuclear Material using Knapsack Problem and Bayesian Probability Models

This talk introduces a framework for screening cargo containers for nuclear material at security stations throughout the United States using knapsack problem, reliability, and Bayesian probability models. The approach investigates how to define a system alarm given a set of screening devices, and hence, designs and analyzes next-generation security system architectures. Containers that yield a systems alarm undergo secondary screening, where more effective and intrusive screening devices are used to further examine containers for nuclear and radiological material. It is assumed that there is a budget for performing secondary screening on containers that yield a systems alarm. This talk explores the relationships and tradeoffs between prescreening, secondary screening costs, and the efficacy of radiation detectors.


Tom McIlvain, DHS

Title: U.S. Strategy to Detect Smuggled Nuclear Materials: The Global Nuclear Detection Architecture

This short presentation summarizes the U.S. strategic outlook and framework on how to reduce the risk of nuclear smuggling and nuclear terrorism, including by exploiting nuclear detection. Global seaports are a key smuggling pathway node for application of nuclear scanning, but they are not the only venue for detection, nor necessarily the most important. Modeling global strategies for detection and interdiction will also be introduced. Eric Tollar (SAIC) will talk in more detail on detection modeling on 11/19.


David Morton, Mechanical Engineering, University of Texas

(joint work with Ned Dimitrov, Dennis Michalopoulos, Mike Nehme, Elmira Popova, Erich Schneider and Greg Thoreson)

Title: Interdicting Smuggled Nuclear Material

We develop a stochastic interdiction model on a transportation network consisting of two adversaries: a smuggler of illicit nuclear material and an interdictor who installs radiation detectors, subject to a budget constraint. The interdictor installs detectors under an uncertain threat scenario, which specifies the smuggler's origin and destination, the nature of the material being smuggled, and the manner in which it is shielded. We describe model variants that differ in how the smuggler selects a path. In all variants, the interdictor's goal is to minimize the probability the smuggler avoids detection. The performance of the detection equipment depends on the material being sensed, geometric attenuation, shielding, cargo and container type, background, time allotted for sensing and a number of other factors. Using a stochastic radiation transport code (MCNPX), we estimate detection probabilities for a specific set of such parameters, and inform the interdiction models with these estimates.


Nishant Pillai, Cargo & Port Security Practice, Unisys

Title: Navigating TWIC Deployment Challenges (Transportation Worker Identification Credential)

TWIC will be fully deployed in all ports by April 15, 2009, yet to-date, numerous challenges have befallen the program. For ports and port/terminal operators to be fully compliant and avoid operational delays in this new security environment, they need to understand the scope of the issues and develop a TWIC deployment/solution strategy that fits in their operational context.


Fred Roberts, DIMACS/DyDAn Director, Rutgers University

Title: Center for Dynamic Data Analysis (DyDAn)

The Center for Dynamic Data Analysis (DyDAn) is one of four Department of Homeland Security (DHS) Centers of Excellence for research on advanced methods for information analysis. Homeland security relies on new data analytic capabilities to ferret out patterns and draw inferences from massive amounts of unstructured data contained in publicly available sources such as newspapers, blogs, geospatial data, images, and audio and video streams. DyDAn researchers are developing new capabilities for absorbing and analyzing the potentially massive amounts of information contained in such media, and they are building the mathematical foundations for computational methods to derive knowledge and understanding from this data while protecting privacy. Homeland security applications of DyDAn's research run the gamut from surveillance for biological events to placing sensors to detect nuclear devices, from border security to protection against invasive species. The broader potential for the technologies developed extends to many other fields that must reconcile large amounts of data from a variety of sources, such as medical informatics, epidemiology, and natural hazard prediction. DyDAn is a collaboration of five university partners and two industry partners: Rutgers University; Princeton University; Rensselaer Polytechnic Institute; Texas Southern University; Texas State University, San Marcos; AT&T Labs; and Alcatel-Lucent, Bell Labs.


Fred Roberts, DIMACS/DyDAn Director, Rutgers University

Title: Sensor Management Problems of Nuclear Detection

The effective use of sensors in nuclear detection requires choosing the right type of sensor, putting it in the right place and activating it at the right times. It also involves interpreting the results of sensor alarms and making decisions that balance various types of risk and uncertainty based on those results, sometimes in real time. We address sensor management for nuclear threat detection by formulating the related problems using precise mathematical language and then developing mathematical sciences tools to solve them. Methods used include applying Bayesian binary and multinomial regression; modifying algorithms from clustering and location theory; applying dynamic programming methods to solve resource allocation problems in sensor management; and investigating new data sampling strategies. Our work concentrates on risk assessment for containers and trucks at borders and seaports including methods to analyze archived data to evaluate trends, get early warning of faulty detectors, and plan manpower and equipment needs. It also concentrates on determining where to locate sensors for special events so as to optimize detection and on detection of devices and materials in moving vehicles or individuals where the problem is to identify the source of an alarm through further measurements and observations.


Captain Dave Scott, Commander, USCG Sector Delaware Bay, Philadelphia, PA

Title: Interagency Cooperation in Port Security Planning, Incident Response & Recovery

In the years immediately following the 9/11 attacks, the Federal government's Port Security Grants were aimed primarily at enhancing the resiliency of our maritime infrastructure. Substantial investments were made to "harden" individual facilities through such efforts as bolstering perimeter fencing and impoundments, improved lighting and video surveillance capability, and more robust deterrents (i.e. the "gates guards, guns and gadgets" approach to port security).

More recently, we've shifted our approach to funding projects that have port wide security impacts, e.g. multi-agency intelligence fusion centers, and improved interagency training and communications. In particular, we've begun to focus on an all hazards approach to incident response, as well at the need to have a coordinated post-incident recovery plan.

I will talk about the evolution of the Coast Guard's port security policies from the early emphasis on "resiliency" to today's focus on "post incident response and recovery", and will cite Sector Delaware Bay's recent development of a Strategic Risk Management Plan" as an example of current thinking. I will discuss open source (non classified) findings of our study to help seminar attendees identify possible gaps and problems that may exist in their port communities, with regard to interagency cooperation and post incident response and recovery.


Philip Stroud, Los Alamos National Laboratory

Title: Nuclear Detection Figure of Merit (NDFOM) Performance Algorithms and Collaborative Database

NDFOM is a database of radiation detectors and radiation spectra, combined with a suite of first-principles detection physics and signal processing algorithms to estimate detector performance in a variety of scenarios. Detailed simulation with MCNPX is combined with published data to develop models of the gamma and neutron radiation spectra from materials of interest, with a range of configurations of shielding and intervening material. Further detailed MCNPX simulation is used to evaluate the detector response functions of PVT, NaI, CsI, CZT detector materials, for a range of detector size, thickness, and geometry. Detector signal to noise algorithms, ranging from a simple total count approach to more advanced energy-dependent matched filter techniques are evaluated. The database currently holds many commercially available detectors, including pager-sized personal radiation detectors, handheld radioisotope identification detectors, large panel radiation portal monitors, and advanced spectroscopic portal monitors.


Thomas Wakeman, Stevens Institute of Technology

Title: DHS Center for Secure and Resilient Maritime Commerce

The National Center for Secure and Resilient Maritime Commerce (CSR) brings together a unique set of academic institutions and public and private sector partners with diverse expertise and significant experience in developing new knowledge, models, tools, policies and procedures, and education/training methodologies related to global maritime security and coastal safety. The Center partners have expertise in national Marine Transportation System (MTS) policy, ocean engineering, maritime security, marine sciences, satellite and radar remote sensing, marine transportation and logistics, systems engineering, oceanography, computer science, naval architecture, physics, sociology, psychology, US and international law, and economics. The partners have worked together in numerous US and international projects related to the safe, secure and environmentally responsible transit of cargo and passengers via the MTS, as well as the short and long-term impacts of coastal hazards on socio-economic systems, ecosystems and living marine resources. These partnerships, resources and capabilities will be applied to the Center's goals of:

  1. Improving port security and the security of coastal and offshore (Exclusive Economic Zone or EEZ) operations and leveraging security investments to also improve economic performance;
  2. Improving emergency response to events in the maritime domain; and
  3. Improving the resiliency of the MTS, offshore operations, and our nation's coastal environments.

The technologies, systems, and procedures that will emerge from CSR will be transformational. For example, sensors, models and systems under development by the partners for coastal environment observing and forecasting can be converted into a dual-use network that provides Maritime Domain Awareness in the coastal and maritime approaches of the entire Exclusive Economic Zone and extending to the high seas. These and other opportunities for leveraging existing investments to achieve both security and economic transformational progress for the nation and for its maritime businesses will be pursued. CSR will function as a resource for DHS in all areas of maritime security and coastal safety, drawing on the expertise of more than 100 PhD-level professionals, with the geographical and technical diversity that is required in a leading-edge, fast-responding technical asset to DHS and the nation.


Denny Weier, Pacific Northwest National Laboratory

Title: Radiation Detection for DHS Applications and the Radiation Portal Monitor Project

In 2002, installation of radiation portal monitors at US ports of entry commenced and today there are over 1000 portals in everyday operation, scanning for illicit trafficking of nuclear material in over 98% of the cargo coming into the country in containers or trucks. This talk will provide information on the Radiation Portal Monitor Project, a Pacific National Laboratory project funded by the Department of Homeland Security to deploy this equipment. The talk will provide details on the installation process and the issues that have been and continue to be overcome during the deployments. In addition, the next generation of equipment, the Advanced Spectroscopic Portal or ASP will be discussed with focus on the testing that PNNL is currently involved with. Finally some thoughts and strategy for the future will be presented.


Chelsea C. White III, Industrial and Systems Engineering, Georgia Institute of Technology

(joint work with Brian M. Lewis and Alan L. Erera, Georgia Institute of Technology)

Title: Managing Inventory in Global Supply Chains Facing Port-of-Entry Disruption Risks

Ports-of-entry are critical components of the modern international supply chain infrastructure, particularly container seaports. The potential operational and economic impact resulting from their temporary closure is unknown, but is widely believed to be very significant. We investigate one aspect of this potential impact, focusing specifically on the use of supply chain inventory as a risk mitigation strategy for a one supplier, one customer system in which goods are transported through a port subject to temporary closures. Closure likelihood and duration are modeled using a completely observed, two-state, exogenous Markov chain. Order lead times are dependent on the status of the port, including potential congestion backlogs of unprocessed work. We develop an infinite-horizon, periodic-review inventory control model for determining optimal average cost ordering policies under linear ordering costs with backlogged demand, and a method for evaluating policy cost. When congestion is negligible, the optimal policy is state invariant. In the more realistic case of non-negligible congestion, this result no longer holds. For the scenarios considered, numerical results indicate that operating margins may decrease 10% for reasonable-length port closures, margins may be eliminated completely without contingency plans, and expected holding and penalty costs may increase 20% for anticipated increases in port utilization.


Posters


T. Altiok, B. Melamed, A. Almaz, A. Ghafoori, CAIT/DIMACS Laboratory for Port Security, Rutgers University

Title: Risk Analysis of the Vessel Traffic in the Delaware River and Bay Area

The project covers the analysis of vessel traffic in the Delaware Channel, including current practices in handling dangerous cargo vessels and vessel delays at Delaware Bay. The economic impact of the vessel activity along the Delaware Channel. Risk analysis and mitigation strategies for safe and efficient traffic management and port operations. Prioritization of Delaware Channel vessel traffic in the course of recovery from a channel-closing high-consequence incident (collision, ramming, grounding, fire, or explosion, stemming from an accident or a terrorist activity).


Jerry Cheng, Rutgers University

Title: Statistical Cluster Detection and Pervasive Surveillance of Nuclear Materials Using Mobile Sensors

Abstract: Pervasive surveillance of nuclear materials provides an effective way to protect against terrorist attack. The advancement of technology makes nuclear detection devices both economic and portable. The GPS (Global Positioning System) is becoming commonly available. It is feasible for the mass production and installation of such devices on taxi cabs in major cities of U.S.

In this work we use spatial multi-cluster classification methods to analyze the signals from the proposed sensor network. Formal statistics tests provide significance against random chance of potential clusters. For illustration, we conduct a simulation study for cluster detection based on several practical scenarios. The results show that the methods can detect clusters accurately and dynamically. Background signals are easily handled with the methods. We can also separate the known harmless signals and treat them as the background.


Ned Dimitrov, University of Texas at Austin

Title: Interdiction, Search and Forensics for Smuggled Nuclear Material

We compare and contrast several models of smuggler movement through the transportation network. Focusing on informed Markovian smugglers, we propose methods of computing interdiction strategies, search plans, and post-event forensics.


Noam Goldberg, RUTCOR, Rutgers University

Title: Dynamic Programming for Efficient Container Inspection Policies

We consider the problem of optimally combining any given set of stochastically independent diagnostic tests into an inspection system that maximizes detection subject to a budget constraint. Our inspection policies can be represented as mixtures of decision trees. We present a dynamic programming algorithm that enumerates all efficient, undominated, inspection policies. Our approach directly solves the bicriterion optimization problem, allowing for sensitivity analysis.

Joint work with: Endre Boros, RUTCOR, Rutgers University; Paul Kantor,SCILS, Rutgers University, SCILS; and Jonathan Word, RUTCOR, Rutgers University


Alexander Gutfraind, Cornell University

Title: Optimal Interdiction of Unreactive Markovian Evaders

The network interdiction problem is a discrete optimization problem on a network that arises in a wide variety of areas including military logistics, infectious disease control, and counter-terrorism. In the classical formulation one is given a weighted network G(N,E) and two agents the ``evader'' and the ``interdictor''. The interdictor's task is to find a set of b edges whose removal would maximally interfere with the motion of the evader on the network.

Our work is motivated by cases in which the evader has only incomplete information about the network, or lacks planning time or computational power. For example, when the authorities set up roadblocks to catch bank robbers, the criminals do not know all the roadblocks' locations or the best path to use for their escape. We introduce a model of network interdiction in which the motion of one or more evaders is described by Markov processes on G and the evaders are assumed not to react to interdiction decisions. The objective is to find an interdiction node or edge set, of size at most b, that maximizes the probability of the evaders captured by that set.

We prove that like the classical approach this interdiction formulation is NP-hard. But unlike the classical problem the objective function is submodular and the solution can be approximated within 1-1/e using a greedy algorithm. Additionally we exploit submodularity to introduce a ``priority'' (or ``lazy'') evaluation strategy that speeds up the greedy algorithm by orders of magnitude. Taken together the results bring closer the goal of finding realistic solutions to the interdiction problem on global-scale networks.

(This is joint work with A. Hagberg and F. Pan in Los Alamos National Lab)


Birnur Ízbas, Bogaziši University

Title: Risk Analysis of the Transit Vessel Traffic in the Strait of Istanbul

In this study, we have analyzed safety risks pertaining to transit vessel maritime traffic in the Strait of Istanbul and proposed ways to mitigate them without causing additional vessel delays. Actual vessel traffic operations and data were studied and a simulation model was developed to mimic the operation with its geographical and weather dynamics. Safety risk analysis was performed by incorporating a probabilistic accident risk model into the simulation model. A scenario analysis was carried out to study the behavior of the accident risks, with respect to changes in the surrounding geographical, meteorological and traffic conditions. This analysis allowed us to investigate the impact of various factors on the risk profile of the Strait. These factors included vessel arrival rates, scheduling pilotage and overtaking policies, and local traffic conditions.

(This is joint work with Tayfur Altiok of Rutgers University.)


Hady Salloum, Stevens Institute of Technology

Title: National Center for Secure and Resilient Maritime Commerce and Coastal Environments (CSR)


Previous: Program
Workshop Index
DIMACS Homepage
Contacting the Center
Document last modified on November 18, 2008.