Graduate Student Research Seminar Day ‑ Nov 22, 2023

1300-1315

1315-1340

1340-1405

1405-1415

Break

1415-1440

1440-1505

1505-1530

Abstracts

Vaccine allocation and supply chain management to minimize epidemiological impacts

Jacob Locke, MASc student

Vaccines are a vital tool in healthcare and one of the most cost-effective ways to curb the spread and impact of disease. Used preventively, such as in childhood and seasonal vaccination campaigns, they help prevent outbreaks from ever occurring and have successfully eradicated or nearly eradicated diseases like smallpox, polio, and measles. Used re-actively in the event of an outbreak, they help prevent the worse symptoms of infectious diseases, while also lowering the transmission rate, both of which help prevent out-breaks from overwhelming healthcare systems and save lives. In the event of an outbreak, vaccines and other resources are often limited, as was seen in early stages of the mass vaccine drive for the COVID-19 pandemic. It was important to make the best use of limited vaccines to both minimize costs and maximize the societal benefit by reducing and controlling the extent of the outbreak.

My research focuses on modeling the course of novel epidemic diseases and the impact different

vaccination strategies will have on the final size of an outbreak, while also modeling the supply chain network to estimate costs. I’ve been exploring different operations research techniques to find the optimal vaccination strategy that balances the final size of the outbreak against the cost of the strategy.

My presentation will cover the formulation of a model that models the dynamics of the epidemic and sourcing and shipping costs for a vaccination strategy and explain why it has proven difficult to solve. Then, discuss three of the different solving algorithms we’ve tried. Covered how they work, comparing their performance, and the pros and cons of each method. Finally, the presentation will briefly discuss some of the results of a case study based on an Ontario during the COVID-19 pandemic.

________________________________________________________________________________

Comparative analysis of node and hub structures in Physical Internet logistics: Evaluating the impact on performance and efficiency

Ruohao Chen, MASc student

The Physical Internet (PI) has emerged as a revolutionary concept with the potential to optimize logistics systems. This thesis article presents a comparative analysis of two logistic models: the PI logistic model and the Traditional logistic model. The PI logistic model incorporates consolidative hubs between nodes, while the Traditional logistic model lacks such hubs for consolidation. The aim of this research is to investigate how the structure of nodes and hubs affects the performance of these logistic models.

Four different structural configurations are explored as designed problems: tree shape, square shape, line shape, and cluster type. By adjusting the location of nodes and hubs, and manipulating the distances between them, the effects of these parameter changes on the models are observed. The objective is to determine which model is more effective for different shapes and types of data. three logistic case studies are conducted in Eastern Canada, Mexico, and Europe to further examine the applicability of the models in real-world scenarios.

To evaluate the logistic models' performance, various metrics such as cost efficiency, inventory levels, and resource utilization are analyzed. These factors provide insights into the advantages and limitations of each model in different structural settings. The research contributes to a deeper understanding of how the physical arrangement of nodes and hubs impacts logistics operations' overall efficiency and effectiveness.

The findings of this research provide valuable insights for decision-makers in optimizing logistics systems by guiding the selection of appropriate models and configurations based on specific requirements, aiming to advance logistics practices.

________________________________________________________________________________

Impact of greenhouse gas emissions on the performance of Physical Internet

Xinyu (Emily) Wang, MASc student

As transportation emission policies become increasingly stringent on a national and global scale, it is crucial for industries to prepare for upcoming taxes and penalties. This research will focus on examining the impact of greenhouse gas emissions on the Physical Internet. Specifically, it aims to investigate how the flow between hubs is affected when considering carbon emissions constraints. The study integrates the concept of consolidation into a MILP (Mixed-Integer Linear Programming) model, which incorporates a Poisson arrival rate, cost parameters, inventory parameters, and emission parameters. It considers constraints related to flow balance, inventory, and capacities, on an example case study taken from the literature.

The analysis presented a multi-tier emission penalty system to closely reflect real-world scenarios. By comparing one-way hub to hub models with and without emissions constraints and exploring the implications in a two-way model, the research provides valuable insights for developing static models and dynamic programs to simulate the effects. Understanding these dynamics is crucial for industry preparedness and sustainable transportation planning, facilitating informed decision-making and the development of strategies to address evolving emissions policies, taxes, and penalties.

________________________________________________________________________________

Optimizing acute stroke treatment process: Insights from micro-tasks durations in a prospective observational time study

Gizem Koca, PhD Candidate

Stroke is a devastating disease and one of the leading causes of disability. However, it is treatable with Thrombolysis (alteplase or Tenecteplase) and Endovascular Thrombectomy (EVT). Both treatments can be given to patients together or individually, depending on their contraindications. Rapid treatment is crucial for acute stroke patients since 1.9 million neurons are lost every minute without treatment. The current literature provides various strategies to optimize the treatment process, but identifying micro-tasks and respective durations may provide further insight into delays and help to propose more improvement strategies. This study aims to identify the micro-tasks involved in acute stroke treatment and measure their durations. A two-phase observational time study was conducted between November 1, 2021 and June 15, 2022. All micro-tasks were identified and categorized in Phase I, and data analysis was completed in Phase II. The micro-task durations were analyzed based on comparisons between daytime and off-hours and between stroke neurologists and non-stroke neurologists. The study found that longer durations of micro-tasks were associated with non-stroke neurologists and/or off-hours operations, and improvement strategies were proposed based on the results.

________________________________________________________________________________

A comprehensive analysis of major maritime accidents in Canada: Identifying causal factors through Systems-Theoretic Accident Models

Shahrokh Bairami-Khankandi, MASc student

In this research, a thorough examination of significant maritime accidents in Canada is conducted, utilizing the Causal Analysis based on Systems Theory (CAST) method to pinpoint prevailing causal factors. The study meticulously analyzes maritime accident reports from the Canadian Transportation Safety Board, offering a systematic exploration of changes in causality over time, across various ship types, and different accident categories. The research questions are centered on the systemic origins of these accidents, with a specific emphasis on inadequate control or feedback failures from controlled entities. The analysis is grounded in the Hierarchical Control Structure (HCS), a conceptual diagram that underscores feedback control loops within a functional system. The results indicate that the most recurrent causal factors are not merely specific to the ship and accident type, but are also deeply embedded in systemic issues. The data, obtained through a consistent and rigorous application of the CAST method, provide valuable insights for academics, policymakers, and industry stakeholders. The findings highlight the necessity for improved safety protocols and strategies for risk reduction in the maritime sector. Moreover, the research underscores the significance of comprehensive investigations and the broad dissemination of their results to effectively address safety concerns in the global marine industry.

________________________________________________________________________________

An Integer Programming approach to employee schedule modeling

Emma Leeshanok, MASc student

Arguably one of the most common and versatile problem archetypes in Operations Research, the scheduling problem has been tackled by countless researchers over the years. This project looks to apply integer programming to the problem while also developing an Excel scheduling tool for the client, Canadian Blood Services. The purpose of this model is to provide the client with daily schedules that is optimized to meet their donation yields across Canada while also being able to prioritize aspects such as the number of employees receiving full time hours or minimizing a clinic's labor hours per unit. The program can be broken down into three distinct parts: a database, a solution picker and a shift/break scheduling problem. The latter two use integer programming to pick viable solutions based on user input and translate it into a schedule which is then outputted back to the user. Although the format of the schedule template and the model parameters are based on the client's needs, the overall concept can be applied to a wide variety of scheduling problems.

________________________________________________________________________________