Risk assessment to mobilize assets

To know where to deploy assets from the Superhero Network, it is important to understand which are the potential hotspots for an outbreak of the virus.

Currently, a lot of people are trying to build AI models to predict the outbreak of the virus. The problem with these models is that the data being used is highly unreliable and the β (uncertainty/unknown factors) is too high.

Consider this scenario where AI model results can be very misleading

• AI model predicts that social distancing can contain the virus by April 28.

• Spread of the virus gets contained with effective social distancing triggered by the lockdown.

• Once the lockdown eases, borders are eased and people from other hotspots enter the COVID-19 contained areas.

• This changes the β significantly.

• From this point, the virus' spread patterns are no longer predictable.

  • This is because there is not enough data to model the influx of people through porous borders.

  • It then becomes mostly biased and based on guesswork than actionable intelligence.

Identify hotspots, mobilize resources

1. We feel that being prepared is very important.

2. COVID-19 is known to increase mortality if the infected person has a high morbidity owing to diseases like diabetes, hypertension, cardiac problems, past infections of TB, respiratory illnesses, etc.

3. In such vulnerable people, the disease progression is faster and leads to pneumonia and multiple organ failure.

4. To identify areas with a high number of vulnerable population, we need to create a dashboard that shows the risk level of various panchayats in the state.

Risk level of various panchayats based on existing anonymised data

1. Anonymise and collect existing data about health conditions.

2. A scoring system that weighs the risk level of a person based on the existing health condition can be created.

   This scoring system will give us a total risk level of a user with X% for diabetes, Y% for TB, Z% for cardiac history, etc.

3. The processed anonymised data is then split across panchayats to identify the historically vulnerable panchayats.

4. By overlaying this dashboard with real-time consultation data through telemedicine, medical emergency calls (for ambulances), medicine delivery, etc., we will be able to get an accurate picture of which vulnerable panchayats are moving into a higher threat level.

5. Alerts are created when risk levels of a panchayat increases.

   1. These alerts are sent across to the district/panchayat administration.

   2. The Superhero Network managers are notified to increase volunteer recruitment in these hotspots.

   3. This will ensure that resources are allocated to the places that needs it the most.

Risk visualisation dashboard - Classifies panchayats based on a combination of anonymised historical data and realtime requests for help in the Superhero Network.

When the risk level changes, alerts are shown. These alerts can also be send as messages or notifications assets in the Superhero Network who can take actions.