My research concentrates on using operations research models to study applied problems in risk and decision analysis, with a particular emphasis on large-scale disruptions. I have applied these models to Hurricane Sandy, the Japanese tsunami, and the Deepwater Horizon oil spill. The goal of my modeling efforts (which include simulation, optimization, dynamic programming, and multi-attribute value theory) is to help decision makers allocate resources better prepare for and respond to disruptions. Many large-scale and small-scale risks deserve to be studied and analyzed. Unfortunately, I do not have plans to advise new students in 2019. If my plans change, I will update this site.


Homeland security and emergency management

As part of this broad effort, Eva Regnier and I recently developed a simulation tool to help the Marine Corps Reserves make better decisions when preparing for hurricanes. The Marine Corps Reserves is using this simulation as part of their training exercises for hurricanes so that they can understand the uncertainty in hurricanes and make better decisions in the face of that uncertainty. In 2017, Anna Prisacari (a former Ph.D. student in Human-Computer Interaction), Sophia Heatherington (an IMSE undergraduate research assistant), and I designed an experiment that compares how users make hurricane preparation decisions without using the simulation and how they make decisions after practicing with the simulation. The students in my engineering economics undergraduate course participated in the experiment. See the preliminary results from this study here.

In partnership with the Iowa Office of Homeland Security and Emergency Management (HSEM), I am working on modeling the interrelationship among capabilities for emergency preparedness. The Federal Emergency Management Agency (FEMA) describes 32 capabilities for emergency preparedness in its National Preparedness Goal. There is a need to understand how these capabilities influence outcomes of primary concern to federal and state governments, such as preventing terrorist attacks and mitigating the consequences from hazards. Capabilities are also related to each other, and focusing on reducing the gap in one capability may help reduce the gap in other capabilities. My goal is to numerically assess the relationships among the capabilities and map capabilities to fundamental outcomes. This goal will be achieved via three steps: (i) creating a network among capabilities within each of the five mission areas; (ii) identifying the fundamental outcomes for each mission area; and (iii) numerically assessing the relationships among the capabilities and how they impact the primary outcomes.

During the past two years, Sri Sritharan has formed an interdisciplinary research group in Hazard Mitigation and Community Resilience (HMCR). The research group involves approximately 30-40 professors. Most of these professors are in engineering, but professors from the Colleges of Design, Business, and Liberal Arts and Sciences are also involved in this research group. As a member in this group, I am active in exploring new multidisciplinary collaboration opportunities with other professors.

I develop and solve models to determine how to optimally allocate resources to prepare and respond to disruptions. A journal article in Annals of Operations Research determines how to allocate resources to respond to and recover from the Deepwater Horizon oil spill. Another article in Risk Analysis presents and solves a model for how to optimally allocate resources to enhance resilience. The article applies this model to recommend how ConEdison should allocate resources to increase the resilience of its electric power sector after Hurricane Sandy. Amro Al Kazimi, an undergraduate research assistant, and I co-authored a paper that determines how much to spend to prepare and respond to multiple disruptions, and we apply the model to preparing for a hurricane (such as Hurricane Katrina) versus an oil spill (such as Deepwater Horizon). Lei Yao, one of my master’s students, and I extended the resource allocation model to preparing for disruptions with deep uncertainty.

Much of my earliest research focused on the economic impacts from disruptions, and two of my most cited papers are: (i) an analysis of the macroeconomic impacts of the 2011 Japanese tsunami and (ii) an evaluation of the economic losses due to a closure of an inland waterway port. You can find copies and links to these papers on my Publications page.


Engineering design and manufacturing

Since arriving at ISU, I have become involved in several engineering design and manufacturing problems. I approach problems in this area by modeling uncertainty in the engineering design and manufacturing processes and then determining the best alternative while considering the uncertainty.

Chao Hu and I are developing an active collaboration in engineering design with a specific focus on designing resilient systems. Designing resilient engineered systems that can sense and withstand adverse events and recover from the effects of the adverse events is increasingly seen as an important goal of engineering design. Our research proposes a value-driven design for resilience (VD2R) framework in order to enable the assessment of system resilience and the optimization of decision variables (or design characteristics) that maximize the value of the system for a firm. My Ph.D. student, Ramin Giahi, extended this work to analyze how a risk-averse firm should consider designing for resilience.

The Center for e-Design has funded Chao and me to research design optimization under long-range uncertainty. Traditional engineering design assumes that customers and the public know what it wants and requirements will not change over time. The new system will operate in a stable environment in which the regulations, technologies, demographics, and usage patterns will not change. However, many engineered systems and products are used in ways that are not originally intended by the designer. This research seeks to identify key parameters in a design simulation that will generate the largest changes in the design parameters. After identifying the key parameters, we will focus on identifying robust and flexible design solutions via an optimization algorithm that account for these long-range uncertainties.

Caroline KrejciGuiping Hu, John Jackman, and I worked with two master’s students to forecast long-term demand for Boeing Aerospace’s aircraft and optimize Boeing’s painting schedule. One of the master’s students, Minxiang Zhang, wrote his thesis as part of this project in which he used probability models to forecast demand for Boeing aircraft 20 years into the future. He also compared different decision-making methods to help Boeing determine if and how to expand its painting capacity while considering the uncertainty in demand.

I am part of research team led by Frank Peters that was recently awarded a grant from the Defense Logistics Agency and the Steel Founders Society of America. The goal of the project is to assess whether casting is as reliable as welding for building structures.


Supply chain risk management

The focus of my research in supply chain risk management stems from an investigation of the impacts of the 2011 Japanese tsunami. I have advised students on extending this work, and I have also pursued new opportunities to focus on supply chain risk identification and mitigation.

I published an article modeling and simulating the supply chain impacts in the automobile industry from the 2011 Japanese tsunami in IIE Transactions. Xue Bai and Andre Fristo, both undergraduate research assistants, worked with me to extend this simulation to include more suppliers and firms (approximately 64 suppliers and firms) that were directly or indirectly impacted the tsunami. Amit Sonar and Arun Vinayak, two master’s students, both published research on supply chain risk as chapters in a book Supply Chain Risk Management: Advanced Tools, Methods, and Developments published by Springer. Amit’s research focused on extending the Wagner-Whitin ordering model to situations with risk and uncertainty. Arun’s book chapter modeled the recovery of firms after a disruption and quantified the impact of consumers returning to a firm after a supply chain disruption.

I am working with Hugh Medal on supply chain risk identification and mitigation for low-volume, high-value (LVHV) supply chains. LVHV manufacturing firms (e.g., aerospace, power plant construction, energy exploitation, shipbuilding) are generally neglected in the supply chain risk management literature, and their problems and solutions are different from high-volume manufacturing (e.g., automobile, fashion). Mike Sherwin, who also manages supply chain quality for a supplier of equipment for nuclear power plants, has been collaborating with us and developing this research for his own company.

Xue Lei, on of my Ph.D. students, extended Mike’s work on supply chain reliability for LVHV firms to model supply chain risk using dynamic fault trees. Dynamic fault trees, which are relatively new in reliability analysis, model the dependency among possible component failures and how these probabilities change over time. Xue’s research applies dynamic fault trees to model supply chain risk for different types of supply chains and under different production scenarios.