I have had the opportunity to work with a tremendous group of undergraduate students (some of which who completed co-terminal programs to complete their B.S. and M.S degrees) at RPI and am looking forward to doing the same at Clemson. I believe that an undergraduate (or Master's) research project is a great learning opportunity for the student(s) to apply what they have learned in their courses to new areas. I really appreciated my own opportunity to participate in a Research Experience for Undergraduates at the Math Department at Lafayette College in 2003. You can find the abstracts of many of the completed projects on this page.
Student: John Wilt
Date: Fall 2017-Spring 2019
Type: Master's Thesis
Abstract: We examine the value of coordination in disrupting transnational illicit trafficking supply chains, specifically large-scale heroin networks. We are interested in assessing the impact of forming various cooperative agreements (like task forces) between law enforcement and intelligence agencies responsible for interdicting different segments of the illicit supply chain. We assess this impact by measuring the improvement to the interdiction efforts that result from different levels of cooperation amongst agencies. Although our case study is focused on heroin trafficking networks into the United States, our framework can be applied to problems in disrupting human trafficking networks where the composition of cooperative agreements and/or tasks forces are not as clear and there is a shorter history of potential interdicting agencies interacting with one another. We provide an overview of the interdiction model used in this analysis, the various methods to account for cooperation between agencies, and insights into the key agencies that should cooperate in their interdiction efforts. Our findings indicate that coordination is critical in networks with distinct smuggling paths since uncoordinated efforts are likely to duplicate interdiction efforts along the same path.
Student: Felipe Ortiz
Date: Spring 2016-Fall 2016
Type: Master's Thesis
Abstract: In this thesis, a simulation is constructed to model the illegal drug supply chain interactions with user networks, law enforcement networks, and rehabilitation networks at the city level with the aim to gain insight as to the effects of state-level government resource allocation. The majority of the policy surrounding the drug markets has been aiming at reducing the supply side of the illegal drug market, where our model is capable of analyzing both policies aimed at reducing the supply side and policies aimed at reducing the demand side. The model built incorporates the street level interactions between the aforementioned entities and tests the effects of changes in government controlled resources. Following the construction and validation of the model, a full factorial experiment was run testing four major government affect variables. These were police budget, police focus (on enforcement or information-gathering), rehabilitation budget, and finally rehabilitation success. Each of the factors had 3 levels, resulting in 81 different experiments varying government resource allocation. Descriptive and inferential analysis was conducted on the results. The analysis offers insights and possible policy changes that could provide value to the war on drugs. These insights include: (1) there is a significant interaction between the rehabilitation budget and the police budget which needs to be considered to attack the demand side of the market, (2) with a non-finite population the relapse rate has a rather low impact on the response variables measured, and, therefore, the return on investment into rehabilitation is similar for a variety of relapse rates, finally (3) focusing resources on gathering information to arrest high-ranking criminals and disrupt higher levels of the supply chain has a smaller impact on the response factors than focusing on street-level enforcement.
Student: Brendan Howell
Date: Fall 2014-Spring 2015
Type: Master's Thesis
Abstract: In this thesis, a new supply chain restoration model is created that aims to model the impact that yielded unmet demand has on customer loyalty in supply chain systems during periods of restoration. When an extreme event such as an earthquake or a hurricane disrupts a supply chain system, diminishing the amount of supply that facilities are able to produce in that supply chain system, certain restoration strategies can be carried out as a means of restoring capacity and ultimately maintaining customer loyalty. Our model incorporates a back-up-facility-purchasing strategy that allows for disrupted facilities to receive temporary supply that can be used for meeting customer demand. Upon creating this new supply chain restoration model and implementing it in AMPL, using CPLEX as the integer programming solver, we offer some policy insights on supply chain restoration that reflect the results gathered in our analysis, including the fact that contracts for back-up facilities are only purchased when the supply chain system intends on exhausting that back-up facility for its total supply capacity. When a supply chain system uses back-up facilities to minimize customer loss, most of the back-up supply that is produced is shipped to the facilities which can then best ship units of supply least expensively to the customers that the model aims to avoid losing. Additionally, when the supply chain system is able to replace lost customers quickly, it is more willing to lose those customers in the system as opposed to carrying out restoration efforts to satisfy their demands. Lastly, especially as it applies to selecting restoration strategies following a large disruption, tradeoffs are made in the attempt to maintain the loyalty of a select group of customers in the system. This typically results in the loss of a larger number of other customers in the system that have lower demand and customer-loss costs associated with them.
Student: Ali Svoboda
Date: Summer 2014-Spring 2015
Type: Undergraduate and Master's Research Project
Abstract: The focus of this project is on the redesign of existing supply chains. A new set of network design problems were studied to explore ways to help companies update their existing facilities while minimizing the impact on current operations. For example, companies may need to update their facilities to accommodate energy-efficient technologies while minimizing the total operating costs of the company during the upgrade period. A key component of this redesign is that the supply chain will remain operational as it is updated. These approaches were examined through the use of network optimization and integer programming. A model was created to represent a company with a national reach trying to reduce their carbon emissions output while still meeting demand. Emissions are reduced in the model through upgrading manufacturing lines to have better energy efficiency; however, during the upgrade period the line’s production is shutdown. This model was implemented in AMPL and analysis was run to help provide new insights into the transitional phase of supply chain redesigns. Due to the nature of this model, there are two conflicting objectives: minimize cost and minimize carbon emissions overages. The tradeoffs between these objectives were examined by varying the penalty cost of emitting a unit of carbon over the set emissions goal. A higher penalty cost places a higher priority on minimizing emissions overages while a lower penalty prioritizes minimizing cost. Analysis was also run to determine whether the company should shift production to another facility or ramp up inventory during a transitional phase of a facility. These results may be beneficial to companies in that if the problems that come with this transition are more understood, companies can redesign their supply chains more efficiently and at a minimum cost. This will ideally make green transitions more feasible and therefore, more likely to be adopted in the future.
Student: Ali Svoboda
Date: Summer 2014-Spring 2015
Type: Undergraduate and Master's Research Project
Abstract: The focus of this project is on the redesign of existing supply chains. A new set of network design problems were studied to explore ways to help companies update their existing facilities while minimizing the impact on current operations. For example, companies may need to update their facilities to accommodate energy-efficient technologies while minimizing the total operating costs of the company during the upgrade period. A key component of this redesign is that the supply chain will remain operational as it is updated. These approaches were examined through the use of network optimization and integer programming. A model was created to represent a company with a national reach trying to reduce their carbon emissions output while still meeting demand. Emissions are reduced in the model through upgrading manufacturing lines to have better energy efficiency; however, during the upgrade period the line’s production is shutdown. This model was implemented in AMPL and analysis was run to help provide new insights into the transitional phase of supply chain redesigns. Due to the nature of this model, there are two conflicting objectives: minimize cost and minimize carbon emissions overages. The tradeoffs between these objectives were examined by varying the penalty cost of emitting a unit of carbon over the set emissions goal. A higher penalty cost places a higher priority on minimizing emissions overages while a lower penalty prioritizes minimizing cost. Analysis was also run to determine whether the company should shift production to another facility or ramp up inventory during a transitional phase of a facility. These results may be beneficial to companies in that if the problems that come with this transition are more understood, companies can redesign their supply chains more efficiently and at a minimum cost. This will ideally make green transitions more feasible and therefore, more likely to be adopted in the future.
Student: Jane Braun, Kevin Toth, and Cheryl Tran
Date: Spring 2015
Type: Master's Research Project
Abstract: The goal of this model is to analyze various options for sensors surrounding Alaska and its waters, and to determine which of these sensors would optimize the safety of the routes taken by vessels in maritime activity in this area. The safety of each arc is determined by the amount of flow on that arc by various ships. The goal of the sensors would be to increase awareness of where each vessel is in the network, along with potential hazards on the paths. Each sensor - types including satellites, UAVs, and AIS sensors - will have varying effects on each arc traveled in the network depending on their functionality and location. Vessels will be determined by their origin and destination, and their paths will be determined by the model with regards to the total safety value. The output of this model is the flow on each of the arcs, and the sensor decisions on whether to install one or not. A budget constraint limits the amount of different types of sensors which can be installed. The network of Alaska that was developed includes major and minor ports in Alaska, along with major continental “destination nodes” such as Iceland, Russia, Asia, and the continental USA. In order to develop various paths through the network, the seas are various other passages were discretized using travel nodes and arcs. An optimization model for sensor location and selection was created to understand the benefits of increasing the available budget and the impact on total safety throughout the region.
Student: Jonathan Holman and Huy Nguyen
Date: Spring 2013 - Spring 2014
Type: Undergraduate Research Project
Abstract: This project introduces the new concept of restoration interdependencies that exist among infrastructures during their restoration efforts after an extreme event. Restoration interdependencies occur whenever a restoration task in one infrastructure is impacted by a restoration task, or lack thereof, in another infrastructure. This work identifies examples of observed restoration interdependencies during the restoration efforts after Hurricane Sandy as reported by major newspapers in the affected areas. A classification scheme for the observed restoration interdependencies is provided which includes five distinct classes: traditional precedence, effectiveness precedence, options precedence, time-sensitive options, and competition for resources. This work provides an overview of these different classes by providing the frequency they were observed, the infrastructures involved with the restoration interdependency, and discussing their potential impact on interdependent infrastructure restoration. Our analysis is important since it provides a new understanding of how the restoration efforts of infrastructures are linked across systems and motivates the need for potential information-sharing in interdependent infrastructure restoration. This project has further created large-scale optimization models and algorithms to determine the loss in restoration effectiveness from decentralized decision-making in interdependent infrastructure restoration and the mitigation of this loss resulting from information-sharing.
Student: Faith Michael
Date: Spring 2010 - Spring 2011
Type: Undergraduate and Master's Research Project
Abstract: The focus of this project is on providing optimization models in order to mitigate the effects of a hurricane on a power infrastructure system by pre-positioning generators at demand points in the system. These models involve two-stages of decision-making: the first stage decisions determine the locations of generators within the system while the second stage decisions focus on the operations of the system after the hurricane. We examine two-stage decision-making models under both a stochastic programming and robust optimization framework. Further, we examine the trade-off curves (or efficient frontiers) for considering the 'average' performance versus the worst-case performance after the event. Our models are tested on a realistic infrastructure system of a county in North Carolina.
Project: Building the disruption scenario for pharmaceutical manufacturing in Puerto Rico after Hurricane Maria.
Student: Jacob Szwarcberg (B.S. RPI IME '21)
Project: Identifying the correct granularity for models for Arctic oil spill response planning
Student: Kelly Steele (B.S. RPI IME '19)
Project: Measuring the impact of improved coordination in disrupting illicit trafficking supply chains.
Student: John Wilt (B.S. RPI IME '18, M.S. RPI IME '19)
Project: Scheduling the recovery of a single supplier in a multi-echelon assembly supply chain.
Student: Zach Shearin (B.S. RPI IME '18)
Project: Resource allocation for decreasing risk in offshore oil and gas development.
Student: Conor Wood (B.S., M.S. RPI IME '17)
Project: Understanding risks from extreme events in multi-echelon assembly supply chains
Students: Lauren Croft (B.S. RPI IME '17) and Shamus Wheeler (B.S., M.S. RPI IME '17)
Project: On the impact of extreme events on multi-echelon assembly supply chains
Student: Benjamin Robinson (B.S. RPI IME '17)
Project: Decision analysis for timing interdictions and surveillance operations against a criminal
Students: Victoria Tong (B.S. RPI IME '16), Wingyan Chan (B.S. RPI IME '16), Benjamin Byeon (B.S. RPI IME '17), and Tianchen Sun (B.S. RPI IME '17)
Project: Examining the impact of cyber outages on infrastructure restoration
Student: Joyce Liu (B.S. RPI IME '16)
Project: Modeling of Arctic oil spill response capabilities
Student: Matai Blacklock (B.S. RPI IME '16)
Project: Implications of unmet demand in the context of supply chain restoration
Student: Brendan Howell (B.S., M.S. RPI IME '15)
Project: Supply chain redesign problems for integrating next-generation manufacturing capabilities
Student: Ali Svoboda (B.S., M.S. RPI IME '15)
Project: Determining the optimal installation of sensors in the Arctic to improve maritime situational awareness
Students: Jane Braun (B.S., M.S. RPI IME '15), Kevin Toth (B.S., M.S. RPI IME '15), and Cheryl Tran (B.S., M.S. RPI IME '15)
Project: Simulating the impact of law enforcement efforts against illegal trafficking operations
Students: Ian Halter (B.S. RPI IME '16), Matthew Macchi (B.S. RPI IME '15), and Tyler Williams (B.S. RPI IME '15)
Project: Decision analysis for determining the optimal conditions for interdicting or targeting a criminal
Student: Victoria Tong (B.S. RPI IME '16)
Project: Examining restoration interdependencies in after-action reports from Hurricane Sandy
Students: Eric Acchitelli (B.S. RPI IME '16), Jared Jensen (B.S. RPI IME '15), and Tyler Williams (B.S. RPI IME '15)
Project: Reverse logistics for humanitarian supply chains after extreme events
Students: Leonardo Heringer da Silva (Exchange student), Arjun Kancharana (B.S. RPI IME '15), and Joshua Moriarty (B.S. RPI IME '15)
Project: Operations Research modeling of sports training
Student: Tyler Williams (B.S. RPI IME '15)
Project: Financial applications and tutorials of Operations Research Methods
Student: Matthew Macchi (B.S. RPI IME '15)
Project: Identifying, classifying, and modeling restoration interdependencies after Hurricane Sandy
Students: Jonathan Holman (B.S. RPI Math '14) and Huy Nguyen (B.S. RPI IME '14)
Project: Mathematical models to measure effectiveness in mission-oriented programs
Student: Lucky Cho (B.S. RPI Math, M.S. RPI IME '14)
Project: The importance of trust in humanitarian logistics
Student: Melissa Licato (B.S. RPI IME '13)
Project: Robust optimization and stochastic programming for mitigating the effects of hurricanes on infrastructure systems
Student: Faith Michael (B.S., M.S. RPI IME '11)
Project: A network-based integer programming formulation for optimizing the location of generators in preparation for a hurricane
Student: Molly Margolis (B.S. RPI IME '11)