Projects

Here are the projects for the fall 2021 semester. You can also see the projects for fall 2019 and fall 2020.

Contents

• Projects
• Mentor assignments
• Pitch recordings
• Intellectual property
• Project details

Projects

After a team finishes Assignment 2, their web site will be linked here.

Mentor assignments

Students, please let me know when a meeting time has been established and I will post it here.

Pitch recordings

• Mon Aug 23
• Wed Aug 25

Intellectual property

Clients marked with ^IP in the list below have indicated that they are sensitive to issues of intellectual property and/or confidentiality. You can read their response about particular concerns in their application. Such projects will require students to review and sign a legal agreement to participate in the project. Students are not required to participate in IP-sensitive projects, and their grade is in no way affected by their choice of whether to engage in such a project. The agreement has been written by the UNC Office of Technology Commercialization (OTC) with this particular situation (and students’ needs) in mind. For more information, see the participation guidelines. You can also view the participation agreement itself. (It’s short.) For questions, please contact Peter Liao <liaopb@unc.edu>, Commercialization Manager at UNC OTC, who will also attend class on Fri Aug 20 to discuss these issues.

Project details

Project titles (provided by the client) and lead client names are given in pitch order below, with links to the client-submitted application and any additional materials provided by the client. Project summaries (provided by clients) are further below.

The applications are responses to this form, which you can see for more details on what questions are being asked, etc. You can page through the application, but please don’t submit an application.

Projects marked ^iOS include iOS development work, so at least one team member must have access to MacOS. Projects marked ^iOS-only are exclusively iOS development work, so all team members must have access to MacOS.

A. Amanda Brumfield - Honors Carolina Database For Resumes & Beyond

Honors Carolina needs a browser-based tool that would give students an easy way to sort through a large volume of document samples to find the ones most relevant and helpful to them. The initial use case would be a resume database, but this could grow to any type of document/file, such as cover letters and personal statements.

B. Rebecca Clark, Matt Verber - ^iOS Sweepstat Mobile App

Instructor note: this project has an iOS development component, so at least one team member will need a Mac.

We are developing a mobile app, for both Android and Apple devices, that can be used to interface with a portable potentiostat (the instrument used to make electrochemical measurements) via Bluetooth. We have worked previously with teams from this course and have made great progress so far! So we are looking to continue working with a new team to wrap up some finishing touches on previous and add additional features. We are mainly looking to improve/finish the GUI, add automated analysis, add the ability to upload/view cloud data, and evaluation/remediation of instrumental noise.

C. Amy Maddox - Strength^n, for meaningful connections

people crave meaningful connection to something larger than themselves. being part of an organization has the potential to satisfy that need, but often we have limited connections. Strength^n (strength to the power of “n”) is a randomization algorithm that allows an administrator to make combinations from a list of participants, at random, and avoiding previous combinations. participants can be classified, and combinations can be make within or among classifications. the administrator tool works well, but there is no user interface other than an email announcing your new combination. the goal of this project is to create a user interface to facilitate scheduling with your new combination, visualizing your growing network of meaningful connections, and maybe even getting conversation sparks. come and help make this simple yet powerful community-building tool even better!

D. Daniel Toben, Ildar Sagdejev - Litter Map GIS Rendering

Plastic pollution is increasing exponentially, and it is a really big problem. The future of humanity involves cleaning up our waste. Help an environmentalist that has cleaned up over a million and a half pieces of plastic pollution add a data science feature to their website under development! Build an engine in any language to pull data from Open Street Maps, perform a logic, and output data in any format, then render that output in ArcGIS/Google as a Heat Map. This heat map will eventually be a predictive overlay on the the website to show locations that likely have a littering/plastic pollution problem. Your work will be fully incorporated into the website, and you will be credited as you please!

A bonus if your output can be a ready to go raster object that can be added to our Leaflet JS widget immediately.

The website we are developing will be a one-stop-shop for people interested in plastic pollution mapping and cleanup. In short, it lets people report locations that have litter/pollution, Local municipalities are automatically updated, and a global map is formed. People who want to volunteer can share their cleanups on social media, and predictive litter mapping can become possible. We believe that the heart of environmentalism needs to be developing a caring heart for the planet, and a primary way of doing that is through acts of service.

The project here specifically is to collect key:value pairs from Open Street maps, perform a logic (can provide or you can leave it flexible), use proximity analysis in GIS, to output data is a raster object. imagine like a weather maps radar or a GIS heat map layer. It will be a semi-transparent layer that will go over our global map that helps to predict where pollution is. Look forward to your help!

E. Willie Barron, Ryan McHargue, Harvey Duperier - ^{IP iOS-only} WAVE

Instructor note: this project requires signing an IP agreement.

Instructor note: this is an iOS-only project, so all team members will need a Mac.

WAVE is a nightlife-centered network of technologies that optimize both the bar owner and bar goer experiences. This project will focus on the development of the bar owner-facing side of the network, integrating data and analytics to augment bars’ marketing efforts.

F. Matthew Simon - Collect SMART

Regional and state level data commonly exists but high-quality primary community level data is hard to find. Community level data is needed to unmask the disparities within a community. Reliable high-quality primary data is needed to help assess community needs. Validated community sampling methods are needed to collect accurate, high quality primary data that don’t require extensive expertise in population sampling, geographic information systems and costly software and computer equipment. Many local health departments do not have the resources to collect this type of data and lack sampling expertise. Thus, Collect SMART would help aid health departments and other community-based organization in collecting high quality data at a low cost.

Collect SMART allows for the rapid collection of community-level data. It improves the timeliness of data collection while reducing the cost of purchasing mapping/data collection software and equipment. The beta version of the ap and website have been tested in 3500 interviews in five states. It has been proven to be an efficient and cost-effective solution for obtaining high quality, representative community health data. The software can be used by hospitals, local health departments, emergency managers or any other community-based organization to collect information about the status of their community. Collect SMART has two parts; the project management dashboard website and the mobile app

The web application uses U.S. Census api to determine the population in an area and implement a random sample with probability proportional to size. Once a sample is drawn, the application allows the project manager to push the sample out to the app, directing survey teams to the appropriate address to conduct an interview. The web application then allows a project manager to monitor the progress of surveyors and analyze incoming data from the field. The project dashboards allows the manager to see the number of surveys completed by each team and by sampling unit. A separate desktop application can be used to design the questionnaire and export an .xml file which can be uploaded through the web application and pushed out to the app.

The mobile app was built on the open-source framework of Epi Info, a free statistical software/platform developed by the CDC. The mobile app adds a mapping module which allows the project manager to upload spatial data, such as census blocks or predetermined survey locations. The mobile app takes questionnaires built in the desktop version of Epi Info, allows users to navigate to the survey locations and collect responses. It is compatible with Android tablets and phones. With electronic devices data can be collected easily and with fewer errors. To ensure data security, a cloud-based database is used for survey responses. Along with the features of collecting data, the app will be able to communicate with other teams in the field and the project manager.

G. Jennifer Stone - ^IP BigWords

Instructor note: this project requires signing an IP agreement.

Billions of dollars are spent to remediate the ‘skills gaps’ in our education systems. The gaps are preventable through daily read aloud practices. But parents either don’t know or are overwhelmed by the demands of modern life. Physicians want parents to read more and hand them paper book logs to fill out. Parents need a tool to help focus caregiving teams on the one thing science shows babies’ brains need most: words.

BigWords is a FitBit for words. It provides caregiving teams a communication platform and feedback on the quantity and quality of words babies hear in storybook reading. Additionally, BigWords provides personalized word-based book recommendations to keep the word collections growing. BigWords grows big brains. BigWords is growing the innovators of the future, one word at a time.

H. Vitaly Radsky, Rocky Moon - Easy Access Counselor Portal

Easy Access is an early stage social good start-up building college search and advising tools to help low-income and first generation students access college. The product that we are currently building is a caseload management tool for school counselors working with high school juniors and seniors. Based on our experience in schools, high school counselors often lack any form of caseload management. They do not track meetings with students, and often don’t have a good idea of what each student is planning to do after high school. With caseloads of 200-400 students, many students fall through the cracks.

Over the past two years we have worked with CS for Social Good, Summer of Code, and two semesters of CS523 to build a prototype counselor portal that allows high school counselors to manage their students. We are looking to build on top of the existing product by adding features including: the ability for multiple counselors to edit and view a cohort of students, a web-based college search interface, and refining the data upload and download process.

I. Peter Halpin - Online Peer Learning for Intro Stats

This project is about extending web apps built using R-shiny to include collaborative features, such as those available via convergence.io (a JS library). I am pretty wedded to R-shiny for content development, but am flexible in terms of how to “collaborify” the content – convergence seems promising but there may be better options.

J. Dezbee McDaniel, Graham Watkins - ^IP Health Equity Project (saving lives!!)

CliniSpan Health is a company that is helping POC, women, and people in rural areas get access to clinical research. The access to research then creates more efficient medicines for those particular groups. These groups take improper dosage amounts, see inadvertent side effects, and generally react more negatively to some medicines because enough people of that group were not involved in initial testing. Think of how this might affect your mom or grandmother. You can help on the journey to create medicines that make their health states much better! This software project will help to effectively reach and engage those communities of people in order to achieve better healthcare outcomes for them.

K. Brian Hu, Paul Tunison - XAITK: The Explainable AI Toolkit

We are developing an open-source explainable AI toolkit (XAITK), which serves as a collection of resources and tools for explaining and interpreting “black-box” machine learning models. Specifically, we have created the xaitk-saliency Python package, which generates visual saliency maps to help reveal what a model pays attention to when it makes its predictions (e.g. for image classification, object detection, etc.)

We have so far developed an API in terms of Python classes and functions. In addition to this, we want to have a web API to expose algorithm configuration and operation for over-the-web scenarios like cloud computing. We envision this web service being containerized to standardize the application environment as well as to be compatible with various cloud frameworks. An additional service that provides a web UI for interacting with this “backend” service, provided as a separate server and container, would also be of use as a demonstration example of both algorithm functionality as well as integrating with the backend service.

You can learn more about the overall effort at the following website: xaitk.org

L. Neal Shah, Gavry Eshet, Autumn Konz - ^IP Personalized Reminiscence Therapy

Instructor note: this project requires signing an IP agreement.

Our mission at CareYaya is to utilize our skills and energy, to make innovations to improve the quality of life for your grandparents and people like them.

Goal of the project is to deliver Personalized Reminiscence Therapy through a mobile app, to create a valuable method of engaging senior citizens, help reduce feelings of isolation and depression, and improve life satisfaction.

If we design it well and store the data from the reminiscences, others in the family may enjoy learning things about their parent or grandparent that they didn’t even know before!

And the memories could be preserved for the future, in a journal electronically that the family can keep and look back on.

A real moonshot goal, depending on your skills and how well we develop this, would be to create a digital story of the person’s life. This digital story can live on even after one day the elderly person passes away, so that the family can have something to remember them by.

M. Brennan Barber - HOLIND: Mobile app to facilitate at home treatment for disabilities

HOLIND is a product designed to capture therapist treatment recommendations and patient progress data.

It eliminates the need to take/receive handwritten instructions for at home therapy and allows the patient to easily record the progress made between appointments in the clinical setting. Often times appointments therapist instructions can be lost in translation, or simply lost! This will be an easy to access platform that captures all data in an accessible way for patients and providers.

O. Penny Gordon-Larsen, Alexia Kelley - Software for Research Impact

In a partnership between Gillings and the Health Sciences Library Informatics team, we propose a demonstration project for generating tools for showcasing UNC’s research impact. We propose to begin with a model for Gillings that could be adapted for broader campus strategies.

For this demonstration project, we are seeking ways to show the impact of the research being conducted by Gillings School of Global Public Health researchers. We have a complex strategy in place with programming already developed and implemented using the SCOPUS literature citation database https://www.scopus.com/home.uri. Our programming works excellently for point-in-time data analysis.

However, we are in great need for developing an API query and results collection system to automate and turbocharge these searches, store the results on Gillings database infrastructure, and allow us to very quickly tap the latest publication information in the SCOPUS database for ongoing impact analyses. The priority for this class project would be, in order:

1. Data retrieval (through API)
2. Data storage (‘locally’)
3. Data analysis/visualization

P. Yichen Yu, JoJo Yang - ^IP Motivation

Instructor note: this project requires signing an IP agreement.

Motivation is a habit tracker app that helps students keep up with their daily habits and homework. It also serves as a platform for students with the same habits to connect with each other and find habit buddies. We have done pre-research, provide all design panels, and also got usability testing results. So all you need is to do the coding part! Most importantly, we have worked with COMP 523 before and we are also students, we understand you the best!

Q. Allen Hurlbert - Avian Diet Database

At the core, this project consists of building a web-based data entry portal. Some of the required features include:

• login/user accounts
• automated QA/QC checks, some of which will interface with external APIs
• a reviewer dashboard to easily see any issues raised by the QA/QC checks
• integration of entered and approved data to an existing database on Carolina Cloud Apps (which can be explored at aviandiet.unc.edu)

R. Taylor Walls - FANzPlay…The Game within a Game!

FANzPlay is an early stage sports and entertainment app that will allow FANs of the most historic rivalries in sports to play against each other LIVE on Game Day to Earn their OWN BRAGGING RIGHTS!

FANS will compete in timed, head to head, sport specific trivia for the opportunity to WIN the gifts, prizes and services being offered by our sponsors AND trash talk! Those sponsors would be given an opportunity to learn data rich analytics on the lifestyles and loyalties of the fans that played the FANzPlay App

FANzPlay will engage fans against the fans they “love to hate” in the competitive match ups to see which fans know the most about the teams,/conference, etc. As we scale, FanzPlay is available for in arena ( HD/ Jumbo Tron Integration )+ out of arena fans!

We understand the pandemic changed the wide world of sports and our focus is to return to the arena or at home with the fans as We Change the way we Play!

If YOU could compete against the FANS you LOVE to hate, would you PLAY?

FANzPlay…The Game within a Game

S. Christopher Gsell - Cafe Reservation System

When cafes get crowded, especially in Chapel Hill and especially with the limited Covid seating, it becomes difficult to find a table. Moreover, many people love to work for long periods of time at a cafe, especially with remote work still prevalent. While it may be ideal for the customer to buy a cup of coffee and sit at the table for hours, it is not ideal for the cafe. This app will solve this issue by allowing customers to reserve tables in advance. If a cafe is particularly busy during a part of the day, they can even charge for these reservations. This app will be a win-win as customers will be assured a table if they reserve it, and cafes will make money from those who before were taking up space.

T. Talpha Harris - Augmented Interactive Education Platform

Ancient Path Adventures is a creative ministry. Our project will combine bible stories with general education topics in an Augmented Interactive Education Platform. Elementary students can enhance their learning journey through augmented books that feature AR components and QR codes. The QR code will lead to the web-app which features: Module quizzes associated with student accounts and leaderboards. Facilitators can track student progress, assign new modules and award students through the platform.

U. Tyler Steelman - ZIP Code Lookup Tool

Researchers attempting to display a survey participant’s congressional district are presently limited in their ability to do so. Asking individual their congressional district is rife with errors as many people do not know the district they live in. One method is to ask a user their ZIP code (which is much more memorable) and then display the corresponding congressional district to the user. However, programming each ZIP code-district pair is time and cost prohibitive. This project will design a web tool that makes this data available as a database that can be queried for use in survey software. Users will provide a spreadsheet of original data (ZIP code-district pairs) that will be hosted and accessible through a URL that sends ZIP code data to the web tool and sends back the ZIP code’s corresponding congressional district.

V. Tessa Joseph-Nicholas - Online Training for CS ULAs

Help us develop a customizable online training for ULAs (Undergraduate Learning Assistants) in the Computer Science department!

W. Scott Maitland - Where’s the Play?

I want to gamify teaching little leaguers where they are supposed to throw the ball based on different game scenarios. Ie. Runners on 1st and Second and ball is hit to you in right field. Where do you throw the ball?

Z. Steven King, Max Hudnell, Daniel Sanchez - VCP Volumetric Capture Processer App

VCP produces volumetric 3D models from multiple video and data sources.

This project will build the public-facing interface to a computer vision volumetric capture project being built by the Reese Innovation Lab.

α. Dimitri Trembath, Paige Hornsby, Laura Stadnik - No More Indecipherable Doctor Handwriting

Lab oversight is a major responsibility for many doctors, particularly pathologists. Despite an increasing reliance on electronic health records for charting patient visits, etc., too much administrative time for physicians is still spent filling out paper forms. Many of these forms, particular in the area of lab administration, are a repetitive process of reviewing/checking boxes followed by signature and dates that attest that the appropriate person has verified certain results. Creating an online version of these forms that can be easily accessed by both laboratory staff and the appropriate physician/administrator will allow for an easier and more seamless route of communication and more robust record keeping required for standard laboratory inspections.

β. Michael Weaver, Delaney Cowart, Erin Carey, Devin Hubbard - ^{IP iOS} Pelvic Floor Smart Wand - Mobile App

Instructor note: this project requires signing an IP agreement.

Instructor note: this project has a probable iOS development component, so at least one team member may need a Mac. It’s possible that the project can be completed without a Mac, but that’s not clear at this point.

Our project entails building a mobile app that will work in tandem with a smart wand to relieve the symptoms of pelvic floor dysfunction. Pelvic floor dysfunction is an extremely prevalent disease, and we hope that the app will help to objectify the therapy process, allowing for treatment to become much more accessible… helping millions of people around the world. This project is submitted by TorchTech Industries, a team that began as a group of seniors in the UNC class of 2021. Thank you for your interest, and we are eager to begin working with your team!

γ. Hari Gopalakrishna - ^{IP iOS} Boomcaster

Instructor note: this project requires signing an IP agreement.

Instructor note: this project has an iOS development component, so at least one team member will need a Mac.

Develop a native mobile application for Boomcaster, a podcast recording web application.