Thesis/Project Final Defense Schedule
Join us as the School of STEM master’s degree candidates present their culminating thesis and project work. The schedule is updated throughout the quarter, check back for new defenses.
View previous quarter schedules
Select a master’s program to navigate to candidates:
Master of Science in Computer Science & Software Engineering
WINTER 2026
Wednesday, February 25
DEBOSHRI DAS
Chair: Dr. Min Chen
Candidate: Master of Science in Computer Science & Software Engineering
11:00 A.M.; Join Deboshri Das’ Online Defense
Project: Blackfoot – Play and Learn: Designing a Game-Based Language Learning Application
Blackfoot (NiitsÃʼpowahsin) is a culturally and historically significant Indigenous language of the Blackfoot Confederacy and a primary medium through which community knowledge, identity, and oral traditions are sustained. As with many Indigenous languages, Blackfoot is experiencing ongoing endangerment, highlighting the need for accessible learning tools and culturally grounded resources that support intergenerational transmission and revitalization.
This paper presents the design and implementation of Blackfoot – Play and Learn, a web-based, game-driven language-learning application that integrates cultural context with structured, practice-oriented activities. The application is organized as a suite of modular learning experiences, including a Blackfoot History module and several interactive activities: Flashcards for incremental vocabulary acquisition, Blackfoot Builder for scaffolded construction of meaning through guided word selection, Voices of the Blackfoot for listening-centered engagement with narrative content and dialect identification, and a Quiz module for rapid recall and self-assessment. A persistent leaderboard provides lightweight gamification to encourage continued engagement and make progress visible across activities. Because Blackfoot is spoken across multiple dialects, the system incorporates dialect awareness throughout its modules rather than assuming a single standardized variety. The system is implemented as a React-based web application using client-side routing and reusable, modular UI components.
The project follows Agile development practices with iterative feature delivery and periodic review feedback. Emphasis was placed on code quality, modular design, reusable utilities, and backend best practices such as normalized answer checking, debounced availability validation, provider-aware reauthentication, and guarded asynchronous data access. The resulting system demonstrates a scalable, maintainable architecture for culturally informed, game-based language learning, and provides a foundation for future adaptive and community-driven extensions.
Friday, March 6
BENJAMIN V. SCHIPUNOV
Chair: Dr. Clark Olson
Candidate: Master of Science in Computer Science & Software Engineering
1:15 P.M.; Join Benjamin V. Schipunov’s Online Defense
Project: AROfficeNav: A Scalable Indoor Mapping and Augmented Reality Navigation System
Indoor navigation has emerged as a growing challenge in large, complex buildings such as university campuses, hospitals, shopping malls, and corporate offices, where traditional signage is not always sufficient to provide consistent, accessible guidance. Outdoor Global Navigation Satellite Systems (GNSS) are unreliable indoors due to signal attenuation and lack of line-of-sight to satellites, while existing positioning systems such as QR-code markers and dedicated infrastructure introduce installation and maintenance overhead that scale poorly over time. These limitations motivate the development of scalable indoor positioning systems that leverage advances in computer vision, sensor fusion, and modern mobile computing. This project presents the design, implementation, and evaluation of an augmented reality-based (AR) indoor building navigation system that integrates 3D spatial modeling to enable scalable, user-friendly indoor guidance.
The system uses a server-backed architecture and a modular reconstruction pipeline to create building-scale 3D maps from phone-captured scans, organized into workspace-based reconstructions to support iterative refinement and long-term data management. For mapping, the pipeline uses learned features and robust pose estimation to produce a sparse 3D reconstruction. For navigation, the system performs real-time visual localization with rolling-window multi-frame matching against a pre-built 2D-3D descriptor database, estimating camera pose using Perspective-n-Point (PnP) with Random Sampling Consensus (RANSAC). A coordinate alignment transform maps reconstructed building coordinates into the user’s live AR session coordinates, enabling stable AR labeling and route visualization. The overall design prioritizes minimal installation and maintenance compared to infrastructure-heavy alternatives such as fiducials and beacons, while supporting multi-floor buildings and scalable deployment.
The project is evaluated in UW Bothell’s Innovation Hall, which involves scanning hallways, open spaces, and staircases. After the scanning process, all rooms, exits, elevators, restrooms, and scan connections are labeled using a label projection system that operates directly on the captured scan images. The scans are then merged into a unified building model, from which a navigation graph is generated and prepared for publication and consumer use. The resulting platform demonstrates the feasibility of scalable, infrastructure-light indoor navigation while highlighting practical challenges in visual positioning under descriptor ambiguity and perceptual aliasing.
Wednesday, March 11
SAMMAN TYATA
Chair: Dr. David Socha
Candidate: Master of Science in Computer Science & Software Engineering
11:00 A.M.; Join Samman Tyata’s Online Defense
Project: Design Decisions and Architectural Rationale in the Luna mHealth Mobile Application: UI Enhancements and Module Page Navigation
This paper presents a critical analysis of design decisions and technical contributions made during the development of the Luna mHealth Mobile application, a Flutter-based mobile health education platform. The work encompasses four primary areas of the user interface: the module homepage, the module reading experience, peer-to-peer module sharing, and a breadcrumb-based navigation system with stack-based history. Each area is evaluated against foundational software engineering principles such as SOLID design, DRY, and Clean Code practices, illustrating how these principles shape architectural decisions in real-world Flutter development. Rather than cataloging all activities, this paper focuses on the reasoning behind significant architectural choices, the iterative thought processes that led to final implementations, and the broader lessons learned about mobile application development in a production environment.
Friday, March 13
PAUL OLADELE
Chair: Dr. David Socha
Candidate: Master of Science in Computer Science & Software Engineering
8:45 A.M.; Join Paul Oladele’s Online Defense
Project: Offline Health Module Sharing in Luna mHealth: A Flutter-Based Peer-to-Peer Approach with Google Nearby Connections
This capstone project presents the design and implementation of a peer-to-peer content sharing system for the Luna mHealth mobile application, an offline-first platform that delivers health education in low-connectivity environments. Luna distributes structured, audio-visual health modules on low-cost smartphones, where internet access cannot be assumed. To support local content dissemination, this work implements a device-to-device transfer system using Google Nearby Connections within a Flutter architecture. The project evaluates prior Bluetooth Low Energy approaches and motivates the selection of a higher-level framework to reduce protocol complexity and improve reliability. The system is modeled using finite state machines to provide predictable connection lifecycles, failure recovery, and multi-peer coordination. A module transfer pipeline with metadata exchange, file streaming, and integrity validation ensures safe ingestion and resharing. The result is a module sharing system that strengthens Luna’s offline distribution model and supports community-driven propagation of health education content.
Master of Science in Cybersecurity Engineering
WINTER 2026
Thursday, March 5
SIDDHA MEHTA
Chair: Dr. Geetha Thamilarasu
Candidate: Master of Science in Cybersecurity Engineering
11:00 A.M.; UW2 (Commons Hall) Room 327
Project: Adaptive Multi-Vector BLE Attack Framework: A Context Aware Approach
Bluetooth Low Energy (BLE) is the standard for modern IoT infrastructure, yet current security tools struggle to test these devices effectively because they lack context. Standard scanners treat every target whether it be a smart lock or a heart monitor as a generic black box target, applying the same random testing strategies to all of them. This “one size fits all” approach routinely misses logic specific vulnerabilities and fails to detect silent failures, where a device’s software crashes while its radio connection stays active.
This project presents “Adaptive Multi-Vector BLE Attack Framework: A Context Aware
Approach”, an automated framework that moves beyond blind fuzzing to intelligent and context-aware testing. The framework uses a Large Language Model (LLM) to plan attacks based on three dimensions: semantic context (inferring what the device is from its data), temporal context (tracking how the device behaves over time), and operational context (mapping its available attack surface). To confirm the presence of vulnerabilities, we introduce a Differential State Engine that compares the device’s state before and after an attack. This allows the system to prove that a flaw exists by detecting persistent changes, rather than relying on simple connection errors.
The framework was tested against a diverse set of simulated ESP32-based targets, which included medical, industrial, and smart home profiles. The testing confirmed that the system was able to successfully identify complex logic-based vulnerabilities, such as authentication bypasses, which are often invisible to context blind tools. These results demonstrate the feasibility and effectiveness of using LLM for planning and decision making with device context to enhance the depth of automated BLE security testing.