Fall 2020 Design I Final Competition
Section X: Team 5
Problem Statement: How can we better mitigate and/or track the spread of COVID-19 at Colorado School of Mines?
Team Members: Kaihl Garcia, Daxton Garner, Eric McKevitt and James Porter
Instructor: Mirna Mattjik
Thank you for your presentation.
1. Opening remark states that each person infected with infect four others, but doesn’t cite a reference or conditions.
2. The financial cost part was presented in the middle of the presentation…it was not clear to me what the purpose of this was, but maybe it was a requirement given by the staff.
3. The assumption of exposure is based upon building location. Is the assumption that if you’re in the same building there’s a risk? Is there a plan to allow a field to include what floor a potential exposed user was on and anonymously ask the same for the person who became infected? Dormitory exposure would cause everyone to request to be tested…is that a good thing or would it cause unnecessary testing?
4. What other applications are there for this sort of system?
5. Are there plans automate exposure notifications instead of having to sift through a list?
6. Would it help to include fields for those who (a) have already recovered from Covid-19 and include their sickness interval? (b) Received a dose of vaccine? (c) Had been tested for antibodies? I can see benefit for providing health care institutions with relevant data to assess re-infection rates and efficacy.
7. The graphics were awesome, however tracking people via GPS becomes sketchy when they go indoors and other buildings are nearby. Did you resolve location in another way?
Thank you, I liked your project a lot. What was your biggest challenge and what would you do different? – Pete
We appreciate your interest in our project. As a team, we have formulated responses to answer your questions to the best of our abilities:
1. This statistic is according to the Imperial Covid-19 Response Team, which is a group made up of scientists from the Jameel Institute for Disease and Emergency Analytics and the MRC Centre for Global Infectious Disease Analysis in the Imperial College Faculty of Medicine. The statistic cited is the Effective Reproduction rate of Covid-19, which means that on average, with all factors and conditions considered, four people are infected with Covid-19 per person who contracts the disease. Here is the citation for that statistic:
Imperial.ac.uk. 2020. [online] Available at: [Accessed 18 September 2020]
2. The purpose of presenting the cost analysis was to provide relevant information for when we are actually able to publish our application onto the app store for both iOS and Android. While a cost analysis was a requirement by staff, we intend to continue development and deemed it valuable to discuss the related costs.
3. The assumption of exposure is dually dependent upon which buildings a user has visited and the relation between their measured location and others’. While we have not yet implemented floor-detection within buildings, it will be our plan to do so. That said, it is also valuable to detect which buildings someone has been in because at the very least, each person inside a building would have needed to enter through one of few entrances, and there is a decent probability that a trace of Covid-19 left on surfaces could infect anyone else who has visited the building within the same timeframe. The same idea pertains to dormitory exposure, although it would certainly be valuable to implement an altitudinal system that differentiates floor levels as you have suggested. With respect to testing, if a user tests positive for Covid-19 and reports that they have done so, the residents of their dormitory will be aware anyway, and this understanding could be used in conjunction with data collected in the app to determine who within the dormitory should be tested.
4. Though the intent is for this application to be applied to Covid-19, it also provides utility in preventing exposure to any number of illnesses. In fact, preventing other illnesses will also play a role in mitigating the spread of Covid-19. Someone who is infected with the flu or common cold is more likely to contribute to the spread of both the virus they are infected with as well as others. Additionally, it is possible that Covid-19 remains an issue for quite some time. This application will help contain Covid-19 and other viruses alike for years to come.
5. Indeed. In the future, we would like to establish thresholds that autonomously define exposure with accuracy and send out notifications to users accordingly. That said, we would still like to provide analytics for manual review if it would be so desired by an administrator or the Mines Covid-19 Response Team.
6. These suggestions would certainly help supplement the sophistication of the application. In the future, we plan to implement a system through which users can report having been vaccinated, which would influence the way the algorithm determines values associated with exposure to Covid-19. Additionally, we could add functionality to categorize users who should have antibodies as of low risk to Covid-19. We would determine a sickness interval to be 14 days, and then an antibody interval to be 3 months, as this is the duration students at Mines are asked to wait before continuing testing after having tested positive for Covid-19.
7. The conceptualization of GPS that most people are accustomed to is a certain type of GPS data called “Coarse Location”. This is the type of location displayed in most GPS applications, which require less precision. This means of collection is relatively inaccurate, so instead, we’ve opted to use a different means of collection called “Fine Location” collection. This allows us to collect latitude and longitude values with incredible precision, and it has been our findings through testing that the location data collected reliably associates the correct building with the collected user location.
To answer your question, our biggest challenge was developing such a sophisticated mobile application and web application in a short timeframe. Because of this, the user interfaces are not as aesthetically pleasing as we would like them to be by the time we are ready to publish the application. Given more time, we would both style our applications and provide more back-end functionality to increase the sophistication of our algorithms and data analysis.
Thank you again for your questions, and we hope we have resolved any inquiries.
HI team,
Nice job on developing a working app for this project. Here are a few questions:
1) Who is the admin and once it is all set up what is their job? Do they need to analyze the data in order for the texts to be sent out or is that automated?
2) If someone tests positive, how far back does the app trace their movements in order to contact others about testing?
3) Is it notifying people that were in the same building for x amount of time? It seems like if you were on a different floor you wouldn’t need to be tested, does it take that proximity into account?
Thanks,
Cara Juergensen
Thank you for your remarks. To best address your questions, we have compiled responses to each as a team which we hope will sufficiently resolve any inquiries:
1. The administrator privilege is an abstraction that could be applied to any figure that should have access to additional analytics. Presumably, the Covid-19 Response Team at Mines would be given administrative privileges to more effectively engage in contact tracing. When the application is published, notifications will be automatically sent out to users who are determined to have been in exposure to Covid-19.
2. The application automatically deletes location from its databases two weeks after collection. This is because after two weeks, one’s location provides no utility in contact tracing and can therefore be removed from consideration. The duration of two weeks was derivative of standards specified by the CDC.
3. The application will not only convey a user’s exposure as a result of the amount of time spent at a particular building, but will also indicate whether a user has been in the building at the same time as someone who tested positive. If two users visit the same building at the same time, the interpersonal contact tracing algorithm will log the interaction to the database, whereas if one user visits the same building that a tested positive user has visited, this would indicate potential exposure as well and would therefore be logged. It is also our intent to establish a system which can distinguish between different floors on a building and adjust analysis accordingly.
Thank you again for your interest, and we hope we have sufficiently responded to your questions.
Good mapping visuals in your presentation!
A few questions:
1) When does the racking actually start? At the log in or when symptoms are added?
2) Is Heat mapping followed only at a certain symptom level? What is that level?
3) Do you have any idea what % of Mines personnel would use this app? Are you concerned people would still be concerned about privacy?
4) Just to clarify… You can only tell people they have been in contact with someone with COVID or with symptoms?
Thank you for taking the time to watch our presentation and for leaving some thoughtful questions. As a group we have formulated responses, and we hope they clear up any confusion.
1. When a user first opens the mobile application, they will be prompted to “Allow Location” as seen in the video demonstration, at which point location collection will begin. The user will have the option to disable location collection at any point in Settings if they desire. If a user decides to disable location collection for any reason, they will still be able to see the heatmap, but will not be able to participate in the data collection associated with contact tracing.
2. A user’s submission to the heatmap will be displayed and logged with any amount of symptoms reported. This is because even if a user has only a few mild symptoms, it could potentially indicate Covid-19 and should be paid attention to on some level of analysis. Covid-19 can manifest itself as any combination of symptoms, so our heatmap should reflect that in its functionality.
3. We are optimistic that a large proportion of Mines students would participate in the contact tracing application. Not only does it provide a user with relevant information regarding the state of Covid-19 on campus, but it allows each student to participate in a process that will contribute to a general feeling of safety on campus. Furthermore, the process will be streamlined and automated as we continue development to provide the users with the most convenient user experience. Because of the aforementioned reasons, we believe that a significant proportion of Mines students would use our application. With respect to user privacy, we have established a system to protect anonymity and ensure that minimal data is stored on each user. Each user is assigned an ID number which has no connection to their identity. The admin will be able to see a user’s ID, but not their name or any other identifying factor. If a user has been exposed to Covid-19, the current process allows an admin to notify the user of their exposure, at which point the user will know they should be tested. In the future, this process will be automated. With this system, the user’s personal information is entirely protected. Moreover, all data collected on the user is available to that user. They can use any of the buttons on the mobile user interface to see the data that is collected for their visits to buildings and their connection with other users. The location data is automatically discarded two weeks after its collection, and users will be able to see this if they enable the “Location History” option, which displays all of the collected location data for that user.
4. A user is able to notify an admin anonymously that they have tested positive for Covid-19. The admin will only see the ID of the user who has reported this, but can use the “Individual Connections” button on the web application’s user interface to compile a log of all of the people that the tested positive individual’s account has been in contact with. In reiteration, this process will be automated in the future, but will currently be facilitated by an admin account. Additionally, we intend to automate the process of reporting Covid-19 test results by connecting with the third party company that administers Covid-19 tests at Mines.
Thank you again for your interest in our project, and we hope we have sufficiently answered your questions.
Nice responses. Thank you for your time and engagement this semester. Good Luck