The virus has affected humanity in various ways, be it our economy, our freedom of movement, and the loss of loved ones. Then how do we live on, comfortably, and safely with this virus around?
Even after the lockdown is over, there is a massive possibility that traces of the virus will remain, and it can spread again.
We wanted to bring people back their mobility and keep them safe at the same time. We wanted people to know about their status while they leave their houses.
What it does
CoviFight alerts me about the risks of catching the virus if I have come in contact with an infected person within the past three weeks. It also informs the healthcare system accurately about the spread of infection.
CoviFight also generates a map with hotspots for what places have virus traces, so that people can prevent travelling at these places and authorities can sterilize or lockdown these places efficiently rather than having a complete lockdown of a country.
How we built it
We develop a three-tier app:
• A user's app
• A provider's app
• An official's portal.
While utilizing Bluetooth and GPS of your phone, CoviFight makes sure that the confidentiality of every individual is secured and can not be compromised. Data is encrypted using a secret key, and no one can view it without your permission. It only traces the past data of positively tested patients. This way, CoviFight also meets the GDPR compliance.
By using Geo-fencing and Machine Learning, we predict your chances of catching the infection so that you can take preventive measures.
A provider's app for aggregation points like shops, restaurants, and public transport synchronizes with the nearby user app. This interface is the key to the detection of infection points, be it a stationary workplace or a moving vehicle. If McDonald's installed CoviFight and had an infected customer in the past 15 days, all the customers after the positive tested patient would get alerted, and hence the restaurant can be sterilized.
Only the medical system may update a person's status over the official's portal, and the authenticity of the app is maintained hence preventing false positives or self-reporting, which might lead to falsification of records.
So, with the help of our app, people can move around while being alerted about their status, stay away from the virus, and be free from the worry of their privacy maintenance at the same time.
Take a look at our demo by clicking here
Challenges we ran into
To maintain the authenticity of the predictions and analysis, we were initially in a fix as to how to update a person's status as positive or negative. Then we decided to come up with a three-tier system, and we developed a Doc App or official's portal, which is only accessed by the medical system so that the authenticity is maintained. No one else can manipulate the data.
Accomplishments that we're proud of
• We have made sure that the privacy of every individual is maintained and can not be compromised. The encryption algorithms meet the standards of the leading social networking apps existing in the market.
• CoviFight not only alerts people about their own risks but provides heatmaps of the traces of the virus too. CoviFight also shows what specific restaurant or public transport( like a bus or a train) may be infected precisely.
• We do not need to compare data between people, thus making computation very cheap and exponentially faster and efficient.
Our Journey So Far
• Winners( Runner up) in the #EUvsVirus, a Pan-European Hackathon
Organised by the European Innovation Council to counter COVID-19 pandemic with more than 9k participants and 2000 teams. We stood second in the Real time Communication and Prevention challenge.
• Top 6 finalists of The Global Hack, by Garage48, April 2020
The Global Hack is a hackathon designed to share and rapidly develop ideas for urgently needed solutions in the face of the COVID-19 crisis, and to build resilience post-pandemic, with over 12k participants from 100+ countries. The team developed a mobile application solution for the containment and tracking of this virus. We were in the top 6 teams in the Crisis Response Track.
• We were also in the Top 23 Student Innovators in COVID-19 SAMADHAN MHRD( Ministry of Human Resource Development, India) Mega Online Challenge
What we learned
It has been an enjoyable experience to work with people who have not even met each other before and still successfully develop this amazing app. We learned a lot through the hackathon, from interacting with the mentors and getting their guidance, to develop the app.
What's next for CoviFight
We plan to get this deployed at its earliest so that people may get their safe mobility back. We plan to deploy this on the Play Store and make a version for iOS as soon as we can. We are also in touch with the Indian Government and we might be able to save lives in India also.
The necessities to continue the project:
• Approval from government authorities to implement and track data.
• Participation from Hospitals/government bodies to update the status of a patient so that system can generate realtime alerts and mark hotspots.
• Cloud resources to scale up the project. Currently limited by the free tier of cloud infrastructure.
The value of our solution after the crisis:
• This application can be used for any contagious disease management.
• It can be used in disaster management to understand the right victims and relief reaches all rightful beneficiaries( such as in the case of floods and storms).
• It can be used by Providers such as McDonald's and Public transport systems to implement targeted location-based marketing complying with data collection practices.
What we have done till now.
• Implemented masked identities for users to comply with GDPR and privacy requirements.
• Identification of hotspots in realtime based on the patient status update.
• Fixed bugs in the flow and to make it work E2E.
• Produced a product demo.
New Technology introduces
• Blockchain to manage user identification data(making it immutable), adding the security of ECC digital signature • Moving from NoSQL mogoDB to a hybrid of BlockChain and GraphDB for better analysis and prediction while keeping the user's id secrete on the Graph • Use of Kubernetes for creating multiple threads for gaining concurrency •Adding caching mechanism to mobile devices to handle any kind of network failure.