While life as we know it seems to have stopped due to the COVID-19, it is important to think of how to make it start again when the moment will be right. The challenge governments are facing is how to get out of isolation in a structured way, without risking another crisis. Inspired, amongst others, by the way South Corea tackled the pandemic, our team wants to provide authorities a tool that would make a safe “de-confinement” possible : a smartphone app called DeCov. Our goal from the start was to provide a support for individuals who will slowly have to return to normal life.
For now, we act to escape and fight the virus by following mostly all of information coming from the media, which relate to the public what the government and scientists state. But these informations aren't specific to each user and it is often hard to find the information you need under the enormous amount of data on the internet. What makes this application really useful is that it gives precise and specific information about how the virus is evolving in the world, and how a given person should act regarding the virus based on their actions. Depending on our space location, our age, the job of our parents, brother or sister, and many other factors, we are more or less exposed to this virus hence the necessity of evaluating this factors and to communicate an index that take into account all of them.
- What it does
For each user, DeCov establishes a social contamination index, computed using several risk-factors. If the user has a high index, going out of isolation presents a danger to others. This could be the case if the user was recently tested positive or is closely related to a covid-19 positive person. A low index however would mean that getting out of isolation is safe. This could be the case if they have been staying in isolation for three weeks without showing any symptoms or if they are covid-19 negative. With time as well as the increasing number of immunized citizens, the risk index will progressively decrease for multiple people, thus assuring a smart & controlled way out of isolation.
DeCov aims to be a non-centralized system that will mainly use Bluetooth in order to identify risky social interactions. As a result, each user’s index will be affected by the indexes of the other users he crosses, all without having access to other data. It is important to note that the user will have to provide little personal information : if they agree to, basic health aspects concerning their covid-19 history will be registered by the app. Since the index will be updated in real time, the data used for the computation will be deleted after that is done.
Additionally, a map representing the ratio of infected cases in different countries, regions and even cities would be made available in order for people to not only see the evolution of the pandemic but also to evaluate their own risk to go out, consequently being able to adapt their behaviour.
DeCov takes into account individual factors and thus provides a tailor-made solution to a very global problem. All this while being easy to implement, to manipulate, and using very few personal data. Local authorities could for example set for a particular risk-level under which people are allowed to go outside - citizens not respecting this index would have to stay at home until it is safe again for them to go out.
-How we built it
We planned to create a smartphone application. To do this we created a first prototype of the app which is presented on the video. Furthermore, we also started the development of the application, but, due to time, we were not able to finish. We also started to code the map that gather all the informations concerning the numbers of contaminated to COVID-19 and sum up the situation. Finally, we established a first algorithm that simulates how would the index change in different scenarios.
-Challenges we ran into
The main challenge was the time variable. Having started the project on Friday, and not earlier, we didn't have much time to contact the necessary people. For instance, we could have benefitted from a conversation with an epidemiologist, although we came across very helpful mentors and research papers, to refine the algorithm calculating the index. Furthermore, we have contacted big associations like the WHO, the ECDC and the CDC to be able to establish a direct channel between Covid-19 related data and the deCov app. However, because of the short time and the current pandemic, they weren't able to get back to us yet. Finally, working long-distance is overall harder than being in a room all together working on the project. Despite all of this, we are extremely proud of what we accomplished in this short time, and we would love to develop our project even further.
-Accomplishments that we’re proud of:
First of all, we are very proud of having accepted this challenge and been able to go through with it - our entire team consisting of first & second year bachelor students, being able to come up with what we have is a great accomplishment. The design of the app was very important to us as it is to be used by every member of society : our refined, clear and aesthetic app aims to make people feel safe, which we achieved. Moreover, finding an alternative to GPS tracking in a project centered on person-to-person contamination was not straightforward. Finally we are very proud of all the ideas we brainstormed and brought up.
-What we learned:
As it is out first hackathon, we learned to work in a team on an interesting subject we were not experts in at all. We met many inspiring mentors that gave us useful tips and tools to work and explore scientific subjects we did not know about. Finally, we learned that every one of us can participate in helping the world get out of this crisis, and we were glad to have found an outlet, through this project, to have our ideas heard.
For the moment, what is missing is the function that gives the probability of transmission of the virus, which is used in a lot of computation for the index. We need some machine-learning skills in order to determine this probability but as we are all young student in bachelor’s degree of physics, for the moment we don’t have these skills. Moreover, the main reason that causes a change in the index are the meeting of two people and the test document to the COVID-19. But, as we know, there are other factors that need to be taken into account, and we would need to meet an epidemiologist, an expert on this field who could access some data to help us evaluate these factors. We need to implement the general algorithm that we explained in the main report, and to connect it with the data. Also, the machine-learning algorithm, as explained to us by an expert, needs a lot of data, and that is another problem that we will need to work on.