V

Vaxi-Van

Vaccine on Wheels

V

Vaxi-Van

Vaccine on Wheels

The problem Vaxi-Van solves

The Problem

The primary concern regarding COVID-19 vaccination in India is the lack of adequate resources for those in need. Getting a vaccine slot is just the start, the person has to risk his life by going to vaccination rooms that are congested and over crowded. Apart from this, old and specially-abled people find it difficult to commute to vaccination centres which can lead to those people being left not vaccinated and posing a threat to their as well as others lives.

The Solution

We will be implementing a novel solution in which clustering up of users that have registered on the portal for a vaccine according to their locations keeping in the mind the condition that a minimum certain number of people are present in that cluster and that the van won’t take up more than the time window for the vaccine to deteriorate will be done.

Each cluster will have its own time scheduled for vaccination in which the users have to show up.
The van will visit one or more clusters everyday according to the number of registrations the portal gets.

We'll be implementing the clustering and the time calculation in the following ways-

  1. A threshold-based clustering algorithm that can give us optimal results for our use case.
  2. Google Maps API to calculate distances and time taken by the van for traveling those distances.

Challenges we ran into

Challenges for implementing

While brainstorming for an optimal solution we ran into some challenges which caused a hurdle in our progress-

  1. Deciding upon the optimal clustering algorithm.
  2. Handling outlier cases where clustering is not possible due to either less number of people or more physical distance between users in a cluster that can lead to wastage of vaccine.
  3. Handling isolated clusters.

We did decide upon the clustering algorithm that would be good for our use-case which would be a threshold based clustering algorithm like QT Clustering Algorithm or Agglomerative Clustering through which we can implement clustering on the basis of the parameters defined in the solution proposed above.

Discussion