Bluevid

Inspiration

Currently there are many people who are afraid of going out, being in contact with their loved ones, etc.

Although the COVID19 does forward, it is currently not known how to control a staggered output, without another regrowth.

The central idea of our proposal is to use mobile devices to collect additional information about those people who may have been in close contact with a positive case and therefore be susceptible to having been infected, so that these individuals can be warned and also the extent of the infected population can be estimated based on confirmed positive cases.

Including a key concept such as gamification, since in the infected detection process you not only favor the need to use the app, but you can also promote civic behavior and social responsibility in a pandemic, as well as using a means that instills a game perspective that lowers the tension a bit on the real situation. Gamification, not yet defined in its entirety, may include bonuses or tips to carry out that, in a “prize” way, reduce your level of exposure.

This concept is similar to the TraceTogether application used successfully in Singapore, which consists of an app that records close encounters between users of the application using Bluetooth connections and allows notifications to be sent to users who have been in contact with a positive case once it is identified. this. BLUVID, based on this concept, proposes the following innovations and improvements:

● Record meetings between users of the application on a server (always ensuring adherence to data protection standards). In this way, they could be used as input for more complex AI-based algorithms and epidemiological models that allow a more precise estimate of the probability of infection of individuals.

● Supplement the above information with geolocation data and possibly with social network data or with additional information entered by the user during the registration of the application. The former would help identify "hot spots" where the infection may be spreading faster. The latter would allow segmenting the conclusions by age, social group, etc. If the user wants to offer this data. In addition to the use of non-relational graph databases, oriented to high performance and obtaining first, second, third ... grade relationships. Likewise, for the MVP, an algorithm based exclusively on signal strength can be created and improved through Machine Learning. Likewise, the device model must be sent as other additional data.

● Notify the user of the application in real time of their risk of exposure to the pathogen. The improvement in the estimation of the probability of contagion would warn you not only if you have been in direct contact with a positive case (contact of order 1), but for example, for having had numerous second-order contacts or when entering a ´hot area´, for example, a shopping center that concentrates numerous individuals who have been in contact with a positive. This algorithm (to be defined) can be such in the MVP and take advantage of data collection to improve its modeling thanks to Machine Learning.

To avoid using the mobile phone with your hands, through the use of helmets or the telephone loudspeaker (not recommended), have an acoustic sensor, which can alert the user through its tone and periodicity when its degree of Exposure increases without looking at the screen.

● The data collected by the application and the results of the data analysis could be put at the service of the health authorities to help in decision-making.

● Modification of the user's status (positive, doubtful, negative) only by medical or authorized personnel after identification via app-app.

● Another different privilege level for sending push notifications with the possibility of geographic segmentation, segmentation by degrees of exposure and user status, remembering the 3 states, perhaps 4 (positive, doubtful, negative, expository), although these states could pass to 5, taking into account an immunization time factor, if the health authorities wanted to use this same application for immunity control.

Regarding battery saving and prevention of housing indication, it has been presumed:

• That the scanning frequency f is f = x (unit of time), that in motion it is f = x + n and in rest situation of f = x - n. Possibility of automatic shutdown established time slot or through user input.

• The user has a button switch to turn on / off the scanning and / or activity function.

• Possibility of including devices of other family members to avoid a match between them and even the option to ignore a device after repeated proximity events appears.