Introduction

This class has originally been built based on Cornell’s Software Engineering version Fall 2023.

Course overview

NoteGoals
  • Ideal: To learn how to deliver quality software when the scope is larger and the stakes are higher
  • Realistic:
    • To work on a project that is significantly more complex than previous courses
    • To learn to take on practices and mentalities that should be as close to a real-world as possible
    • Sprinkles of theoretical concepts about software engineering
NoteThemes
  • Engineering principles
  • Management & teamwork
  • Design
  • Construction & maintenance
  • Quality
  • Delivery
  • Professionalism

Example success and failures

SuccessLinux kernel (1991-present)
SuccessSQLite (2000-present)
Failure: Therac-25 (1982)
Failure: Ariane 5 (1996)
Failure: Boeing 737 MAX (2017)
  • IEEE Spectrum: How the Boeing 737 Max Disaster Looks to a Software Developer
  • Engine size issues when move from 737 to 737 Max design
  • Boeing’s solution to its hardware problem was software (Maneuvering Characteristics Augmentation System - MCAS)
  • MCAS automatically pushed aircraft’s nose down when it thinks the angle-of-attack exceeds its limit (too high up).
    • It also uses Elevator Feel Computer to override the pilot’s ability to manually pulling up (explicit design decision)
    • If the angle-of-attack sensors are faulty (and they were for the incidents), MCAS only believes the sensors!

The course project

NoteTeams
  • Teams
    • 4-8 students
    • Ideal number is a team of 6, but depend on scope
  • Check for compatible schedules and work styles
  • Try to include members with both frontend and backend experience (well-rounded team)
  • Students can form complete or partial team on their own
  • For partial teams or students without team
    • Utilize the Discord server to specify that you are looking for more members (which schedules/skills) or if you are looking for a team (provide your schedules/skills)
    • The professor reserve the right to assign individual members (I will most likely only to it for people who can’t find a team)
  • Teams should be formed by the third lecture, where we will discuss the subject of Teamwork and collaboration
NoteProjects
  • Projects from the course instructor
  • Projects from WCU faculty
  • Teams may petition for an external project:
  • Timeline: four 3-week sessions
NoteProject deliverables
  • Session reports
  • Plan
  • Accomplishments, setbacks, discovered work
  • Peer feedback
  • Work log
  • Design documentation
  • Test plans and reports
  • Coverage analysis
  • Requirements
  • Code and code reviews
  • User documentation
NoteSelf introduction
  • Name
  • Preferred programming language (or one you are most proficient with)
  • Preferred operating systems and platforms/libraries (or one you are most familiar with)
  • Anything else?

Available project topics for Fall 2026

NoteWCUPA HR Rammy AI Chatbot Enhancement
  • Client: WCU HR
  • Description: This CSC 402 Capstone project will modernize the HR Rammy AI Chatbot by replacing its current single-model setup with an advanced dual-model framework. To ensure a smooth transition, the project begins by building an isolated staging environment restricted to HR staff, giving the team a safe space to test updates without risking production uptime. To measure success, the team will create a standardized testing benchmark using real HR questions paired with human-approved answers. Using this benchmark, the project will evaluate a dual-model RAG architecture, separating query orchestration from response generation, across both fully local and hybrid setups against the current OpenAI model. Upon proving that a dual-model approach improves performance, privacy, and control- the team will deploy the upgraded system to production and deliver complete process documentation.
Jetson Tracking Sensor Package
  • Client: SME Solutions Inc.
  • Description:
    • The system is centered around an NVIDIA Jetson, which serves as the primary computing and control platform for the vehicle. A camera and other supported sensors connect directly to the Jetson using standard interfaces such as USB, Ethernet, or GPIO. The Jetson collects and processes sensor data, performs object detection and tracking, determines the desired vehicle motion, and generates commands to control the vehicle.
    • The architecture is intentionally modular so that sensors and software components can be added, removed, or replaced as the project evolves. Individual software modules may handle functions such as camera processing, object detection, sensor fusion, Guidance, Navigation, and Control (GNC), and vehicle control, while exchanging data through well-defined software interfaces.
    • The initial prototype will focus on integrating a small set of sensors and demonstrating an autonomous vehicle capable of detecting, tracking, and following a selected object. This provides a foundation that can later be expanded with additional sensors, algorithms, and autonomous behaviors.
Hiring Process Modernization
  • Client: Chester County Department of Human Services
  • Description:
    • The Chester County Department of Human Services plans to update their existing candidate evaluation matrix to modernize and strengthen the county’s hiring process. Building on the functionality of a data driven ranking system, this project will expand the use of a matrix to incorporate scoring criteria, improve usability for hiring managers, streamline documentation, and support more consistent, data‑driven assessments across all interviews. The updated scoring tool will need to align with established Departmental workflows, existing technological stacks, and other to-be-identified administrative and financial constraints.
Crime Prevention and Personal Safety Application
  • Client: Dr. Jaeyong Choi, Criminal Justice Department
  • Description:
    • This project will develop a crime prevention and personal safety application that uses publicly available crime data to provide users with location- and time-based crime risk information. The application will analyze historical crime patterns and other relevant geographic and temporal factors to estimate the relative risk of crime in a selected area and time period.
    • The system will incorporate a crime risk forecasting component that identifies patterns based on factors such as location, time of day, day of the week, and historical crime trends. Rather than attempting to predict individual criminal events, the application will present users with understandable risk levels or safety information that can help them make more informed decisions about their surroundings.
    • The application will also include several personal safety features. Users will be able to activate a loud emergency alarm designed to attract attention in a potentially dangerous situation. The application will provide easily accessible emergency contact information, including options for contacting 911, campus or local law enforcement, and designated personal emergency contacts.
    • An additional safety-information component will provide users with practical crime prevention and personal safety guidance, including situational awareness strategies, information about responding to threatening situations, and general self-defense information.
    • The project will emphasize usability, responsible use of crime data, privacy, and clear communication of uncertainty. The application will be designed so that crime risk estimates are presented as informational tools rather than guarantees of whether a particular location is safe or dangerous. The initial prototype will demonstrate crime-data integration, geographic risk visualization, time-based risk estimation, emergency alarm functionality, emergency contacts, and personal safety resources.
Self-Choice Food Delivery Management System
  • Client: Oxford SILO
  • Description: Help a local nonprofit organization remove barriers to families receiving the food they need. SILO’s food delivery program helps households with limited transportation in our rural community and those who work during the day when traditional food pantries are open by delivers fresh and shelf stable food directly to the front door of recipients in the evening hours.
    • We currently deliver identical bags of food to our recipients and would like to shift to a self choice model and allow recipients to choose the food in their bags each month that match their dietary restrictions and cultural preferences. We are looking to design and launch an app where friends (clients) can create and sign into an account, confirm their address and phone number, see when their next delivery date is (food is delivered monthly) and choose the food they would like in their next delivery. On the organizational side, we would need to be able to see updated contact information to import into our database, as well as a packing list for staff and volunteers when they prepare the bags. Ideally these could be printed on stickers that are attached to the food bags.
Polaris Brand Promotions Operations Database
  • Client: Polaris Brand Promotion
  • Description: The project will involve continuing the work started by the Spring 2026 West Chester University Capstone Course:
    • This project asks the team to modernize a real operations workflow at Polaris Brand Promotions by replacing worksheet-based storage solutions, one for scheduling shifts/events and one for client billing & contractor payroll, with a secure, cloud-based web application backed by a database. You’ll design the data model so shared fields can flow reliably from scheduling into billing/payroll. You will also work on adding more advanced features such as searchable client and staff profiles, single-shift entry forms plus bulk CSV/Excel import, reporting dashboards/exports for operations and finance, automatic daily backup emails of the data as Excel/CSV, and an audit trail that records which logged-in corporate user created or edited each record.
    • Our agency’s main operational functions are performed using two Google Sheets: our BA Shift Scheduling - Master File and our Client Billing & Payroll File. These files do not directly interface and are prone to human error, so we’re looking for an AWS based online platform that is customizable for our needs and eliminates the functionality issues from our current Google Sheet-based system. See attached example files that we will review during our first meeting.

Challenges

What do you like the least about this class
  • I didn’t like the lectures. They felt like a waste of time when I could just be working on the project.
  • The work for our client could feel disconnected at times from the material we learned in class. The material we learned in class was relevant to presentation projects, but the actual development for the client projects could feel less related.
  • The lack of project choices and independent scoring in projects
Additional comments
  • There were a lot of unrealistic expectations for this class. We only had about 12 weeks to work on the project, and the professor expected us to have a lot more done even though we all have 3 – 5 additional classes on top of this one. If this class is only 3 credits, expect only 3 credits worth of work!
  • Don’t expect us to be Clemson students. There were so many references to “Clemson does it this way” or “Clemson students do this” we are West Chester students so only except result that a West Chester student would be able to give.