Search

Mild Cognitive Impairment (MCI) often occurs prior to the more serious condition of dementia and early detection of MCI is an important but challenging task because of its indistinct symptoms. Work has been done developing serious games on mobile devices for MCI detection as opposed to a typical application of serious games for growing and maintaining mental acuity. The serious games WarCAT and Locker record player’s moves made while playing the game to determine their levels of strategy recognition and learning. To be able to demonstrate this, however, requires a large amount of player data. Therefore, it would be beneficial to develop a method of generating synthetic data that could imitate human player data. The area of machine learning (ML) known as Reinforcement Learning (RL) can be applied to creating a large pool of players since it emulates the way humans learn. In RL, if an action in response to a stimulus is followed by a successful reward, the stimulus-action-reward association will be strengthened, and the reward will be recalled with greater likelihood upon later presentation of the same stimulus and action. Like RL in humans, considerable trial and error (training) is often required. In addition, a growing subfield of machine learning known as Deep Reinforcement Learning uses techniques of Reinforcement Learning along with Artificial Neural Networks for function approximation. The purpose of this thesis is to explore the use of Deep RL to learn to play our serious game and achieve gameplay results comparable to the best human performance. From this, we can define baselines which allow us to create bots with various levels of training to emulate individuals at various stages of learning, or by extension, various levels of cognitive decline.

Over generations, engineering faculty qualifications transitioned from practice to scientific theory basis, and engineering schools progressively shifted responsibility for instructing elements of practical design to industry. The result: a progressive erosion of practical design knowledge (design esoterica) in faculties and industry. Tolerancing based on ASME B4.2 Fits is one element of eroded mechanical engineering design esoterica and is now poorly understood in academe and industry. As a result, students and early-career mechanical design engineers are challenged to select appropriate Fit-types, and therefore to apply appropriate tolerances to their designs. Though unquantified, this leads to elevated industry costs associated with manufacturing and service. With thirty years in mechanical engineering design, over twenty years in the application and instruction of geometric dimensioning and tolerancing (GD&T) in industry, and six years instructing GD&T at the University of Manitoba, the author of this study has observed a general absence of tolerancing knowledge in industrial and academic environs. Recognizing the challenges of learning and teaching mechanical design esoterica in general, and dimensional tolerancing in particular, the author proposes a collection of artifacts as a first step in reintroducing, into mechanical engineering design, a functional understanding of tolerances, and tolerance magnitudes. Three visio and visio-haptic artifacts are designed for use in developing a cognizance of the clearance Fit-classes and the micron-scale tolerances associated with them. The artifact designs are based in the mechanical engineering design experience of the researcher. A qualitative self-study accompanies the engineering design study of the artifacts. A rhetorical Voice of a Design Companion is invoked as anecdotalist to convey the researcher’s design thinking, and to explore the evolution thereof. The scope of the engineering design study is limited to ide

Career development is an important aspect of emerging adulthood and the experiences of post-secondary students. Particularly in the modern context, where technological advancement has significantly altered the world of work and will continue to do so, students need to be equipped with the skills to manage their careers and move forward amidst career uncertainty. This study is a grounded theory examination of students’ experiences of career supports integrated into a biosystems engineering design course offered in the first year of a Biosystems Engineering Program at the University of Manitoba. Very little research has been conducted on the integration of career development into the classroom, though there are promising preliminary studies that speak to the potential positive implications for students. Using the systems theory framework of career development as the theoretical framework, this study seeks to better understand how students in a professional engineering program, who are typically thought to be career decided, experience in-class career supports. The literature review in this thesis examines the central focus of this study, the career development of post-secondary students, and covers two cognate areas, emerging adulthood with a focus on vocational and engineering identity development and reflection in experiential learning. Through utilization of constructivist grounded theory, a rigorous, social constructivist approach to the study design and qualitative analyses of participant interviews and student coursework is presented. The study proposes a low-level theoretical model of these experiences, entitled Becoming a Biosystems Engineering Student, which demonstrates the ways high school students interested in math and science are drawn to first-year engineering, where they tend to use deficit language to describe biosystems engineering. Students then progress to biosystems by either stumbling into biosystems engineering or a planned journey to biosystems

This thesis examines Machine Learning (ML) applicable to detecting Mild Cognitive Impairment (MCI) from gameplay data of a Serious Game (SG). The foundation of the work is a newly developed SG called WAR Cognitive Assessment Tool (WarCAT). WarCAT is based on the familiar card game WAR and is played on mobile devices (phones, tablets). WarCAT captures players’ moves during the game to allow inference of cognitive processes of strategy recognition, learning, and memory. This thesis focuses on the potential of detecting MCI by developing a data synthesis model and a classifier model to be applied to WarCAT gameplay. The data synthesis model generates WarCAT gameplay data similar to that of a human, so that it can be used to train and validate the classifier model. The classifier model has the potential to detect MCI from a player’s gameplay data. Machine Learning (ML) methods are used by both models. The classifier uses a deep Artificial Neural Network (ANN) that is first trained on a large labelled dataset to help detect MCI. The data synthesis model generates synthetic data to plausibly emulate a large population of players. The sub-paradigm of ML, namely Reinforcement Learning (RL) is used to generate synthetic data, as it most closely emulates the way humans learn. An RL bot undergoes millions of sessions of trial and error (training), processing millions of gameplay training patterns and achieves results comparable or better than the best human performance. A conjecture is that an RL bot with fewer training opportunities emulates a person with less experience or potentially declining cognitive function. This baseline allows for the creation of bots with the goal to emulate individuals at various stages of learning, or conversely, various levels of cognitive decline. The thesis demonstrates the ML framework necessary to generate and classify different levels of play, which correspond to different levels of players’ cognitive states or abilities.

Mild Cognitive Impairment (MCI) is cognitive decline before one is naturally expected to have age-related cognitive decline but not severe enough that it affects a person’s daily routine or life. MCI is often a precursor to the various forms of dementia before the symptoms are manifest and diagnosed. If MCI could be reliably assessed, the prognosis of more serious decline may be improved. This thesis will discuss the development of a serious game denoted War Cognitive Assessment Tool or WarCAT, with the objective being to aid in the early detection of MCI. This thesis will also overview contributions to papers published, as well as the work done on the previous iteration of WarCAT denoted War Data on Gambling (WarDOG) whose purpose is to help in the study of gambling and gaming addictions.

This thesis covers the design, implementation, and review of a video game designed to assist Canadian Engineering Interns in understanding and contextualizing engineering ethics. This understanding is essential during their professional practice exam and subsequently in their day-to-day lives as engineers. In engineering schools, engineering ethics is traditionally taught either as a philosophical examination of how engineers should act  or as rote learning of the act, by-laws, and code of ethics that govern engineering practice. Most importantly, in the context of undergraduate engineering education, the amount of coverage is limited, and students are all too often focused on what is needed for the test, not mastery of the material for their own understanding. Unlike university courses, playing this game is voluntary, no grades are assigned, and players are expected to game the system by choosing poor responses just to see what will happen to them. Learning occurs through exploring cause and effect relationships, by making ethical choices and experiencing how decisions often have trade-offs or conflicting right answers. To encourage reflection, players were asked to think about the cases, and how they reacted to the unprofessional behaviour of characters in the game, through this reflection process, players are encouraged to grow, understand, and adopt professional behaviours. The research methodology was to create a proof of concept video game featuring five case studies of conflicts that an Engineer or Engineering Intern might reasonably encounter in their professional practice. The game then went through a design review, in which sixteen Professional Engineers and Engineering Interns played the game and reviewed the cases in detail to provide feedback on their realism and identify areas for improvement. Based on the feedback from testers, the concept is sound, addresses a need within the engineering community and merits further research.

This research is directed towards improving patient care within a hospital’s Emergency Department (ED) by providing support for health care administrators and policy decision makers. The work presented is an innovative automated data collection system, thereby freeing ED workers to focus on the most important aspect of their job: patient care. The thesis is organized as follows; patient data from the Winnipeg Regional Health Authority’s Emergency Department Information System is used to provide insight into the patient flow processes within an ED. This process analysis was critical in developing an agent-based model of process flow (patients and staff) within an ED. From this point, emphasis was placed on illustrating the utility of an Agent Based Model (ABM) in the optimization of a mobile Real Time Location System (mRTLS). Current data collection methods used for localization in health care facilities involving human data collection methods often tend to generate inconsistent and ambiguous data. To address these data collection problems, an automated ceiling mounted mRTLS is proposed. The system utilizes mobile Radio Frequency Identification (RFID) and is evaluated using ABM and optimized using metaheuristic genetic and simulated annealing algorithms. A prototypical agent model for an emergency facility was created using AnyLogic simulation software. The emergency department model is outfitted with mRTLS using mobile, ceiling-mounted RFID readers. Path planning and resource provisioning of mobile readers is accomplished using both a multiobjective genetic algorithm and a multiobjective simulated annealing algorithm. The genetic algorithm optimizes initial placement of static readers by finding areas where RFID tags are more frequently found. This generates the input parameters for the second algorithm, a multiobjective simulated annealing algorithm. The simulated annealing algorithm provides the first-cut path for the deployment of mobile readers. These paths are

Women are underrepresented in the Canadian engineering profession at a rate of nearly 10:1 when compared to their male counterparts. This poses a problem for a profession whose goal is to provide innovative and inclusive solutions for all people. In Manitoba, while women are underrepresented in engineering at the same rate as the national average, there is also evidence that some either never enter the profession after graduation or leave at various stages in their career. This qualitative research study examined the experiences of two groups of women in Manitoba, Former Practicing Members and Graduates Never Registered, to understand the factors that both enable and deter their persistence in the engineering profession. Focus group sessions and one-on-one interviews provided an in-depth perspective of factors contributing to women’s career decisions. The findings generally align with the literature-based conceptual framework and indicate that this is a multi-dimensional problem that includes factors such as a need for improved work-life balance, workplace cultural shifts, and confidence building. Implications of these finding include a need to support both new graduates in the licensure process and former members who wish to return to practice.

This thesis discusses the creation of an agent-based model for vehicle traffic that uses a previously developed mathematical car-following micro-simulation model. The agent-based model results are verified against the original model. The agent-based model is then used to explore the effects of inclement weather and autonomous drivers on traffic to explore the suitability of using the agent-based approach to model traffic. The primary environment focused on is a signalized four-way intersection, with an extension to freeways for autonomous vehicle simulation. The model is used to demonstrate how optimizing green light intervals at intersections based on current weather conditions can help to partially restore traffic throughput to normal condition levels. Results from simulations with autonomous vehicles demonstrate that traffic flow steadily increases as these vehicles enter the driving population. In the case of signalized intersections this improvement increases in inclement weather, but increases are flat across different weather conditions for freeways.

This research addresses the design and development of a data-driven Agent-Based Modelling (ABM) framework to simulate transmission and spread of West Nile Virus (WNV) among heterogeneous mobile humans, various bird species, and Culex genus mosquitoes over a geographic region at a province-scale. A diverse variety of topics and techniques regarding the data collection phase is presented, as modelling WNV has many disparate attributes. WNV is a mosquito-borne disease influenced by avian species as their amplifying hosts. A significant amount of data, such as home range, flight speed, WNV competence index, etc., regarding over 150 bird species along with their population estimates and locations in Manitoba, Canada are estimated and assembled. The primary contribution of this thesis is the development and validation of a data-driven Cellular Difference Equation (CDiffE) scheme for adoption in WNV-ABMs or other mosquito-borne disease ABMs. The migration patterns of different bird species, nocturnal biting activities of Culex mosquitoes, daily temperature and rainfall, and land cover impact are incorporated into the CDiffE model. The CDiffE model at its core employs difference equations, which are computationally faster than commonly used differential equation models. The proposed CDiffE model is cellular to capture heterogeneity of various geographical areas. The CDiffE has been rigorously verified and validated. While the whole system is designed from an ABM perspective at a cellular level, it exhibits biologically compatible behaviour at the macro-level scale. The proposed CDiffE demonstrates high accuracy in predicting real-world mosquito population trends and geographical distributions, evidenced by the mosquito trap data from Manitoba. The proposed CDiffE model updates on an hourly step to act as an environment for a comprehensive ABM of WNV spread among peripatetic humans. Such a hybrid ABM is successfully built on top of the proposed CDiffE scheme to study the imp

The purpose of this research is to analyze factors influencing the adoption of heavy axle loading (HAL) on short-line railroads in Canada. The research comprises a series of case studies which characterize selected Canadian short-line railroads. The approach stratifies the industry in terms of the type of ownership and geographic region. It also documents factors influencing the adoption of HAL for each railroad by examining the commodities it hauls (internal motivation) and its network connections to the Class 1 system (external motivation). Where available, infrastructure condition data are also reported. These studies revealed that all of the 31 railroads studied exhibited a medium-high or high overall motivation to adopt HAL. This reflects the nature of the short-line industry in Canada—filling a niche by hauling primarily heavy commodities (thereby providing internal motivation) while relying on Class 1 partners to offer complete services to their customers (thereby providing external motivation).