2026 Management Engineering Capstone Designs

StreetSmart aims to address the challenge of road safety by developing a risk-aware navigation platform that predicts and compares route safety in Toronto using machine learning and geospatial analytics. The system integrates public traffic, road geometry, and collision datasets to model crash risk and generate safer routing recommendations beyond traditional time-based navigation. By automating data processing and analysis, StreetSmart enables drivers to navigate the safest routes while helping cities better understand safety patterns, with the broader goal of promoting safer and more human-centred urban mobility.

Finding your next favourite movie shouldn’t be a chore. Photuris is a modern movie recommendation platform that allows users to import existing ratings and lists, combining them with dynamic movie data to deliver personalized recommendations. Built to support ethical, consent-based data collection for academic research, Photuris provides transparency and user control while enabling long-term research value. The platform delivers an intuitive, engaging experience for movie lovers while creating a reliable, evolving dataset for academic study.

This capstone analyzes the impact of flexibility (shared departmental OR access) on operating room scheduling optimization, to examine how incremental relaxation of constraints affects the total patient wait cost. Using data from Cambridge Memorial Hospital, ORLex delivers an interactive dashboard that enables the comparison of surgical schedules, wait times, and costs across different flexibility levels. By translating scheduling flexibility into measurable outcomes, the dashboard supports informed decision-making and allows the hospital to evaluate flexibility levels that align with its operational priorities.

Cambridge Memorial Hospital's operating rooms maintain numerous surgeon-specific pick-lists for the same procedures, with minor item variations between lists inflating SKU counts and carrying costs. Picktacular is an interactive decision-support dashboard that visualizes variation across pick-lists for a given surgery, identifies potentially interchangeable items through algorithmic labelling, and suggests consolidation opportunities. The tool enables OR staff to compare pick-lists side-by-side and make informed substitution decisions, supporting inventory reduction without prescribing specific changes.

For every hour of body-worn camera footage, Waterloo Regional Police Services spends upwards of six hours manually redacting, removing sensitive information like faces, names, and personal details before releasing it for court. Our AI-powered audio pipeline dramatically reduces this burden by converting noisy body camera audio into clean, structured transcripts. The system automatically identifies speakers, detects radio communications, and extracts personal information for redaction. By automating these tedious tasks, we enable faster, more accurate redaction, getting critical evidence where it needs to be, sooner.

Coordinating volunteers is a time-consuming yet essential process for community building. Many organizations lack the resources or funding to support an automated volunteer coordination system. In collaboration with Engineering Outreach at the University of Waterloo, we designed Shiftly: a smart volunteer selection and scheduling tool that creates an optimized schedule catered to coordinator requirements and volunteer preferences. Shiftly helps organizations reduce manual workloads, run programs more efficiently, and grow social impact by empowering volunteers to find roles that suit them best.

University students can struggle to select courses that align with their interests, degree requirements, and career goals due to fragmented advising resources and complex prerequisite structures. UWGuide addresses this problem by providing a personalized, centralized course recommendation system that analyzes student interests, academic history, resumes, and program requirements. The system validates prerequisites and degree constraints, explains recommendations, and helps students strategically plan flexible, efficient academic pathways while reducing stress, regret, and delayed graduation risk.

Zentra addresses the gap between how financial concepts are taught and how real financial decisions are made. Traditional financial education is often theoretical and disconnected from real-world outcomes. Zentra is a centralized learning platform that immerses students in realistic, scenario-based market simulations where they must make decisions and adapt as conditions change. By combining behavioral finance insights with interactive, gamified learning, Zentra helps users learn through outcomes and feedback, strengthening financial judgment, confidence, and decision-making over time in a low-risk environment.

You rarely notice infrastructure systems, until they fail. From energy storage to HVAC systems, faults are often detected too late, leading to safety risks, and costly maintenance. Anomalyze is an interactive visualization support platform that translates machine learning based fault detection into clear, actionable insights for engineers. Applied to HVAC pumps, Anomalyze visualizes sensor trends, anomaly signals, and failure risk through an interpretable dashboard, enabling early intervention, confident maintenance decisions, and resilient infrastructure.

Campus Connect addresses the challenge of scattered campus event information spread across multiple social platforms, making it hard for students to stay informed and engaged. Our solution is an AI-powered event booking platform that serves as a centralized hub for all university events, club activities, and committee meetings. Students can easily discover, register for, and get personalized recommendations for events, while organizers gain a simple dashboard to post and manage listings; creating a more connected & transparent campus community.

Partner notification (PN) practices for sexually transmitted infections continue to evolve, yet few system-level tools exist to evaluate their impact, partner engagement, and follow-up. This project presents a secure, web-based provider platform and workflow that supports anonymous partner notification and end-to-end contact tracing. The system enables clinicians to manage cases, track notifications, and document follow-up outcomes. By integrating traceability into PN workflows, the solution establishes a foundation for measuring PN impact, identifying gaps, and supporting ongoing improvement of practices over time.

VolunTrack Ontario addresses the challenges high school students face in discovering and managing volunteer opportunities, including information overload, poor tracking, and a lack of motivation to follow through. Existing platforms focus on listings but offer little support beyond discovery. VolunTrack Ontario is a digital platform that helps students find relevant volunteer opportunities and stay engaged through reminders, progress visualization, and personalized milestone recognition. By centralizing discovery and tracking, VolunTrack Ontario supports more consistent and meaningful student participation in volunteering.

BillBoard is a mobile application that helps Canadians discover and track federal legislation through unbiased, plain-language bill summaries. Leveraging modern NLP and our own recommendation algorithm, the platform automatically summarizes, categorizes, and recommends bills based on user interests. Users can search, save, and follow legislation by topic, making parliamentary activity easier to understand and track. By translating complex legal text into accessible, personalized insights, BillBoard addresses gaps in civic awareness and enables more informed democratic participation.

CoachVision is a basketball scouting and analysis tool designed to help varsity and semi-professional teams better understand their performance. It brings together team stats, player trends, opponent insights, and an AI assistant that allows coaches to ask questions in everyday language. Instead of digging through spreadsheets, coaches can quickly see key patterns and insights to support game planning and player development. The goal of CoachVision is to make advanced basketball analytics easy to use, affordable, and accessible for teams with limited resources.

EEVAO addresses the time-intensive and interdisciplinary nature of early rail alignment design, where vertical geometry decisions must balance a wide range of design constraints under tight project timelines. Changes in vertical alignment significantly influence long-term traction energy consumption which can lock operators into decades of higher energy consumption. EEVAO formulates inter-station vertical alignment design as a constrained energy-minimization problem and applies a feasibility-aware genetic algorithm with fast, physics-based train simulation to generate multiple standards-compliant, energy-ranked alignment alternatives. An interactive interface enables engineers to explore this.

NextTerm is a personal academic advisor designed for University of Waterloo Engineering students. It takes your academic history, program requirements, prerequisites, and available course information to help students make more confident and informed course decisions for their next term.

Pipistrel electric aircraft rely on black-box battery State-of-Health (SOH) estimates, offering little visibility into what actually drives battery degradation. This project introduces an interpretable, data-driven SOH modeling system built from per-second flight telemetry, segmented into meaningful operational events. By uncovering how real flight, charging, and idle behaviors affect battery health, the model reveals previously hidden degradation patterns. These insights translate into practical flight and charging recommendations that extend battery life, reduce replacement costs, and improve operational reliability across electric aircraft fleets.

Finding the right information is hard, and it only gets worse as files multiply across more places. Stella is an AI-powered search assistant built for modern work, where documents live on laptops, cloud drives, shared folders, and team tools. Instead of guessing filenames or digging through folders, ask in plain language and Stella searches by meaning, not just keywords. It returns the best source, highlights the exact passage, links you back to the file, and helps you act. Less searching, fewer duplicates, and more time back.