Industrial and Manufacturing Systems Engineering electronic theses and dissertations (MU)
Permanent URI for this collection
The items in this collection are the theses and dissertations written by students of the Department of Industrial and Manufacturing Systems Engineering. Some items may be viewed only by members of the University of Missouri System and/or University of Missouri-Columbia. Click on one of the browse buttons above for a complete listing of the works.
Browse
Recent Submissions
Item Developing a gemba board deployment and assessment system in a multidisciplinary hospital setting(University of Missouri--Columbia, 2024) Tucker, John Alton; Noble, James S.[EMBARGOED UNTIL 12/01/2025] The healthcare industry has struggled for decades to create and sustain a culture of high reliability, specifically within the continual process improvement realm. Implementing methods by which all staff in the healthcare system have an avenue to foster a culture of improvement is needed. One Lean tactic that aids in those efforts seen in other industries are the use of Gemba Boards. A literature review was conducted, and it was found that the use of Gemba Boards was primarily only in pockets of a healthcare organization, mostly nursing. A gap exists that this research aims to fill is establishing a framework that a hospital can undergo when no system of huddling for improvement exists that is standardized in every facet and department within the healthcare system. A framework was developed in two phases, to establish the initial deployment of Gemba Boards, followed by the progression to a sustained cyclical system and program for the Gemba Boards to be assessed for comprehension and return on investment for high reliability efforts. The framework established in this research was piloted in a study within a hospital system that had no formal huddling for improvement or Gemba Boards as a part of their culture. The integration of Boards into every department in the hospital was analyzed as a first attempt to examine how all facets of a healthcare system comprehend and execute the tool at the same time. The objective through statistical analysis is to advance the confidence in the industry that though uniquely different, with the use of the framework developed, deploying Gemba Boards and the philosophy behind them can produce results in any type of healthcare department and environment. Additionally, analysis will be conducted to gain insight into how the model functions when not just one area or pocket of the hospital is deploying this system of huddling, but the entire organization concurrently when pushed and supported by leadership.Item Power plant vibration monitoring using wavelet feature extraction and functional design of experiments(University of Missouri--Columbia, 2024) Oguejiofor, Benjamin Nwakile; Seo, KangwonIn a nuclear plant power generation, analysis of vibration signal constitutes an integral part of predictive maintenance for rotating equipment such as pumps, motors, turbine generators, etc. Vibration signals are continuously monitored via sensors and thresholds for alarms maybe set up to identify equipment malfunction. Improved methods for decomposition and analysis of power plant vibration signals using wavelet feature extraction and functional design of experiment (FDOE) have not been sufficiently investigated for applicability in analyzing these signals for better detection of equipment faults. Chapter 1 introduces the general concepts and methods to be applied in our research study. In Chapter 2, we present the application of discrete wavelet transform (DWT) to decompose a reactor coolant pump vibration signal into frequency sub-bands and the generation of a number of features from some statistics. A principal component analysis (PCA) is used to reduce the large set of variables into a few principal components which can be applied in future monitoring of normal vibration signals. From the insights gained using PCA, Chapter 3 studies the linear discriminant analysis (LDA) to simulated vibration signals, to distinguish between normal and abnormal signals. In Chapter 4, we apply the functional principal component analysis (FPCA) in characterizing the vibration signals generated under several different levels of an environmental factor, a flow rate, associated with a condensate pump. An FDOE is applied with the target vibration curve and used to obtain an optimal flow rate. The obtained flow rate was found comparable to the theoretical pump curve best efficiency point (BEP) and recommended for use for optimal pump performance and reliability. In Chapter 5, we perform an extensive review of literature on FDOE and its applications. A standard FDOE framework was shown and demonstrated the five basic steps to be applied when using JMP Pro 17. Chapter 6 provided overall conclusions and suggestions for future work.Item Predictive analytics within the collegiate wrestling recruitment process(University of Missouri--Columbia, 2024) Mocco, Peyton James; Noble, James S.In this study, we dive into the world of sports analytics, specifically focusing on wrestling. By harnessing data-driven insights, we aim to revolutionize student-athlete recruitment and development. The research focuses on Folkstyle wrestling at the high school and collegiate levels, where performance data is collected and analyzed. A new prediction model offers a fresh perspective on identifying promising wrestlers. Through rigorous statistical analysis, the model uncovered key factors that correlate with success on the mat. These insights empower coaches, recruiters, and student-athletes alike, providing a competitive edge in talent acquisition. However, the journey does not end here. We acknowledge the limitations of the study--namely, its applicability to other wrestling styles such as Freestyle and Greco-Roman. As we move forward, we encourage fellow researchers to build upon the foundation, expanding data collection and refining predictive models. In conclusion, this thesis bridges the gap between sports and data science, opening doors to transformative practices in wrestling. This study enables us to redefine how champions are discovered and nurtured.Item Multi-objective decision-making in solid waste management including social sustainability consideration(University of Missouri--Columbia, 2024) Gutierrez Lopez, Jenny Pilar; McGarvey, Ronald; Noble, JamesWaste management is a critical sector that needs to coordinate its activities with outcomes that impact society. Multi-criteria decision-making methods for waste management have been widely considered using environmental and economic criteria. With the development of new social regulations and concerns, sustainable waste management needs to additionally target socially acceptable practices. Despite the need to aid solid waste management decision makers in contemplating the three pillars of sustainability, limited inclusion of social impacts has been found in the multi-objective decision-making literature. Therefore, one can observe the importance of understanding the local context in which the waste management system is located, and the essentiality of community consultation to recognize potential challenges and improvements to solid waste management systems. Consequently, the involvement of stakeholders is crucial during the quantification process of social indicators. The purposes of this study are threefold: (1) Develop an interview and literature-based framework for the quantification of social metrics in waste management and their inclusion in single-objective optimization problems; (2) Provide a solution approach for the multi-objective problem in waste management that considers the three pillars of sustainability obtaining a set of pareto optimal solutions; (3) Offer analysis of the tradeoffs of three pillars of sustainability for policy development in waste management.Item AI-driven techniques for enhanced efficiency in psychiatry clinical scheduling(University of Missouri--Columbia, 2023) Kasaie Sharifi, Seyed Alireza; Rajendran, Suchithra[EMBARGOED UNTIL 12/1/2024] In recent times, psychiatry clinics are constantly facing late patient arrivals. Patient unpunctuality significantly affects the use of resources and patient waiting times, among other quality indicators. Consequences of tolerating early and late arrivals include disruption of the timetable and lengthier wait times for incoming patients, which may lead to patient dissatisfaction and physician fatigue. Considering these problems, psychiatric clinics strive to improve their services. Therefore, this research suggests three phases for enhancing psychiatric services using predictive and prescriptive analytics. In the first phase, four machine learning models, including multinomial logistic regression, decision tree, random forest, and artificial neural network, are developed to predict patient arrival patterns and aid efficient scheduling accurately. These models are analyzed using the explainable synthetic intelligence approach and the Shapley additive explanations model, promoting comprehension and trust in our algorithmic results. The results show that the random forest algorithm can predict the patient's unpunctuality with an accuracy of over 85 percent. Also, the travel distance, appointment lead time, patient's age, Body Mass Index (BMI), and specific mental diagnoses are significant factors affecting the patient's unpunctuality. In the second phase, new appointment systems are developed to effectively distribute patient needs during the clinic session to enhance resource utilization and patient satisfaction. The proposed appointment systems can address the patient's unpunctuality, preference, and availability. To do this, 21 different scheduling rules (3 benchmark rules + 18 proposed rules) have been assessed regarding their average total cost under different scenarios. The HSBGDM-1 rule emerged as balanced for clinic operations, proficiently managing physician time but occasionally causing overtime variations. Increased patient delays often heightened physician idle times, especially under IBVST and VBFST rules. Hybrid rules, like the HSBGDM series, adapted well, enhancing patient wait times and managing unscheduled patients. Conversely, REPDM scheduling might extend waits, affecting patient satisfaction. Systems focusing on new appointments can increase physician idle times due to unpredictability. While accommodating unscheduled patients boosts service quality, it may also cause disruptions. In the final phase, a genetic algorithm is utilized to determine the optimal patient-to- physician ratio for minimizing the average total cost in a psychiatric clinic across varied scenarios. The sensitivity analysis revealed a profound understanding of how different patient-to-physician ratios affect outcomes across diverse situations, particularly regarding unpunctuality rates. Notably, an 80:2 ratio demonstrated consistent effectiveness in most scenarios, suggesting its suitability as a benchmark for healthcare scheduling. However, specific unpunctuality rates influenced optimal ratios differently, highlighting a potential non-linear relationship between unpunctuality and ideal proportions.