LetsStayAtHome

Inspiration

Hello, my name is Nicolas Cepeda. In the last weeks a big part of my family was in contact with the corona virus with different severity levels (mild till hospitalisation).

What does the app do?

The #LetsStayAtHome App is designed to help people:

Manage their quarantine
Based on all the data collected calculate the likelihood that the person has been exposed to the virus
Inform about the availability of close by hospitals
Spread official news and help detect fake news
How to act in case of Emergency

Try the app in your phone

Running the code

Clone the app repository

cd app
npm install
npm run start

Risk Modelling

Data Collection

The App collects following data at onboarding:

General data (age, kanton, etc)
Past Health Conditions relevant to Covid
Questions regarding the quality of the quarantine
Symptom check

And prompts the user everyday to do a self check containing following information:

Symptom check
Questions regarding the quality of the quarantine

Algorithm

With all this data the app calculates the likelihood that the user as been exposed to the virus and generates a recommendation (stay at home, etc).

Current Implementation

The current model is based on simple linear algebra and needs to be recalibrated.

ML-based implementation

The anonymized data from the users are being collected in a centralised fashion.

Once enough data has been collected (specially interesting are users that became COVID+ or show symptoms of infection) an ML model is going to be trained to better predict the likelihood of infection.

Resources:

App https://stayathome-chi.now.sh/
Slides: https://docs.google.com/presentation/d/1FSCYgUBLNA0WpQUyPhGrY3s7z-ccpaeFFv6y1i_MFMg/edit?usp=sharing
Source Code App https://bitbucket.org/cepedanicolas/stayathome/src
Source Code Landing Page https://bitbucket.org/cepedanicolas/stayathomelandingpage/src
2 Minutes Video Pitch https://www.youtube.com/watch?v=zQwjiFWyZH0&feature=youtu.be

Problems to Solve

The app focuses in the following problems:

Quarantine management Help manage self quarantine better.
Hospital Help triage for more efficient resources management.
Economic Help reduce the economic impact.
Information Help spread official information and protect from Fake News
Data Anonymized Data Collection for better projection and models

Product Roadmap

Current - v0.7-beta

Try this version on your phone

Onboarding
Symptom Checkup
Simple Risk Modelling
Daily BAG News & Recommendations
Help in case of Emergency
Fake News Detection
Review all text & models
Improved Risk Modelling
Anonymised Data Collection
Landing Page
I18N

Next - v0.7-alpha

App Store Release required

Push Notifications
Push Notifications
Google Maps integration
Canton Overview: Show by Kanton the percentage of people that are green. (Charting Technology from https://devpost.com/software/covid19-live-tracker)
Bluetooth Usage to detected proximity to other people (Seems this project is already doing this: https://devpost.com/software/80_corona-contact-tracker_coronow).
QR-Code based immunity certificate (through scanning of a QR code the person can prove he is Corona Immune and is allowed to go out.)