High-Level Project Summary
Covaware is a mobile solution that aims to consolidate people around the global problem by simplifying the non-intuitive processes around the pandemic. We collect a lot of data in real time, analyze it with our ML model, predict the risks and present it to the user in an accessible way (the most frequent human needs in one click) so that everyone can easily follow the complex mathematical predictions. We use in our predictions the recommendations that we have made to other people (game theory) and the personal characteristics of the user to build a really smart system.
Link to Project "Demo"
Link to Final Project
Detailed Project Description
please see : our presentation
Features
Our first feature is to display and notify the current state of danger around the user. By scanning the user's location, we can warn the user about the danger in time with a notification, which can always alert the user.
With the help of the hazard heat map, the user can always take a safe route to his destination. Thanks to the accuracy of 1 km our map gives free maneuvering space to the user.
A distinctive feature is the feature that is called by the button on the bottom right. It allows the user to select the desired public place among the most popular types of these places. After selecting a place, we display the available places nearby with handy markers that show the risk of visiting this place at the moment.
If the user chooses a convenient place to visit, we'll tell the user at what time it's best to visit that place with minimal risk of getting infected by the virus in the form of a list. In the list we display the most accessible hours for visiting on weekends and weekdays.
Hazard visualization
In order for all users to feel comfortable using our application, we have opted for a simplified system for displaying the dangers, rather than complicated percentages. Our danger system is divided into 5 elements: minimal, weak, medium, high, very high with corresponding colors from green shades to red shades.
Our advantages
The first and indisputable advantage is the number of sources we process and run through our ML model to create our predictions, such as COVID-19 cases, Vaccination policies, Health system policies, Containment and closure policies, Lockdown restrictions, Population density, Public places in the area, Attendance of the place. We also use personalized characteristics in our formula that we get when a user logs into our app, such as age, chronic diseases, and whether the person has been vaccinated.
The second but not least is the approach in which we use some rules from Game theory in our predictions. We take into account our prior recommendations to evenly distribute users across the general area. This approach avoids the problem of Similar predictions.
Also, our main feature, which is related to the recommendation of the place and time of the visit, allows the user to orient to his plans much easier and accordingly to protect himself from the virus.
How it works
The central point of our application is a formula for predicting the risks of being infected with a coronavirus. The Auto Arima model is used to understand, on a time-series basis, the contribution of COVID-19 cases, Vaccination policies, Health system policies, Containment and closure policies, Lockdown restrictions, Population density, Public places in the area, Hours of attendance of the place.
Thus, we can predict the number of new cases in a certain area of the city and, accordingly, the danger of visiting places, in addition knowing the number of public places in a certain area. Once we have an initial risk, we use personalized information and coefficients from scientific studies to refine the risk for that particular user.
Using the open api from Google, we learn the traffic situation in specific public places to display their risk level and recommend the best time to visit them.
By using evenly distributed predictions for all users of our system, we make a final prediction that takes into account the predictions that we have made to other users.
Space Agency Data
We use JHU COVID-19 Dashboard—Center for systems science and engineering at JHU
(https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6) for daily getting COVID-19 cases and vaccination data by a USA county. These are all parameters for our prediction model.
https://github.com/CSSEGISandData/COVID-19
We use Gridded Population of the World (https://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-density-rev11) to get population density with accuracy down to 1km. This parameter is in our prediction model. We have found couple of papers which say that the spread rate of SARS-CoV-2 is strongly associated with population density:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7665678/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8253654/
We use Covid-19 Government Response Tracker (https://github.com/OxCGRT/covid-policy-tracker/blob/master/documentation/codebook.md) This dataset contains information The dataset contains 23 indicators and a miscellaneous notes field organised into five groups C - containment and closure policies, E - economic policies, H - health system policies, V - vaccination policies, M - miscellaneous policies by regions.
We use COVID-19 Vaccinations in the United States,County https://data.cdc.gov/Vaccinations/COVID-19-Vaccinations-in-the-United-States-County/8xkx-amqh for getting daily vacination data.
Population density. For our prediction system we use conclusions and coefficients from this articles:
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7665678/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8253654/
Relationship between age and COVID-19. For our prediction system we use conclusions and coefficients from this articles:
https://www.nature.com/articles/s41598-021-97711-8
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7173682/
Relation between diseases and COVID-19. For our prediction system we use conclusions and coefficients from this articles:
https://www.cdc.gov/coronavirus/2019-ncov/science/science-briefs/underlying-evidence-table.html
https://www.cdc.gov/flu/vaccines-work/vaccineeffect.htm
https://www.cdc.gov/coronavirus/2019-ncov/science/science-briefs/underlying-evidence-table.html
https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/people-with-medical-conditions.html
We have found that the weather doesn’t have a relationship with COVID-19, so we haven’t used it. https://www.tandfonline.com/doi/full/10.1080/10962247.2020.1823763
The paper that has inspired us to consider different local characteristics:
https://academic.oup.com/jpubhealth/article/43/3/455/6082830
Hackathon Journey
Of course, apart from the under-sleeping hours, our team had a great time this hackathon. We were faced with a really serious question, which, in our joint opinion, we solved with dignity.
The global crisis of the new virus hasn't stopped and every person should understand it, we must not just survive it, but also develop a strategy for the same trials in the future. We all understand from the past years that the best way to stop the virus is to vaccinate and isolate ourselves. But people don't always have the ability to self-isolate, so we don't call for breaking quarantine measures, but we want to minimize the risk if necessary, with one application with all the necessary functionality.
People often mistakenly believe that global problems can be solved in a single day and with one good decision, but global crises such as the COVID-19 pandemic require daily, collaborative work by people around the world, even in such trivial things as choosing a time to go to the store or the gym.
Our app, with the help of new technology, offers the opportunity for everyone to take small steps toward a great future!
References
https://gisanddata.maps.arcgis.com/apps/opsdashboard/index.html#/bda7594740fd40299423467b48e9ecf6
https://github.com/CSSEGISandData/COVID-19
https://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-density-rev11
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7665678/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8253654/
https://github.com/OxCGRT/covid-policy-tracker/blob/master/documentation/codebook.md
https://data.cdc.gov/Vaccinations/COVID-19-Vaccinations-in-the-United-States-County/8xkx-amqh
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7665678/
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8253654/
https://www.nature.com/articles/s41598-021-97711-8
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7173682/
https://www.cdc.gov/coronavirus/2019-ncov/science/science-briefs/underlying-evidence-table.html
https://www.cdc.gov/flu/vaccines-work/vaccineeffect.htm
https://www.cdc.gov/coronavirus/2019-ncov/science/science-briefs/underlying-evidence-table.html
https://www.cdc.gov/coronavirus/2019-ncov/need-extra-precautions/people-with-medical-conditions.html
https://www.tandfonline.com/doi/full/10.1080/10962247.2020.1823763
https://academic.oup.com/jpubhealth/article/43/3/455/6082830
https://developers.google.com/maps/documentation/places/web-service/overview
Tags
#risk #map #places #stayhome
Global Judging
This project has been submitted for consideration during the Judging process.