Reading Time: 3 Minutes

The COVID-19 pandemic has created a huge, unprecedented demand for medical equipment and supplies. Hospitals and healthcare units are operating at full capacity. Projecting and managing medical resources is a tough task.

Addressing this issue, Prof P Sunthar, Department of Chemical Engineering, IIT Bombay has developed a web tool for ‘Short-term projection of COVID-19 medical resources and inventory’ in collaboration with JNCASR and IISc. 

Motivation and Problem Solving

The medical inventory projection web application came out of discussions among Prof Sunthar and his colleagues in Bengaluru. They initially considered developing a mobile app that can help the government gather data from the ground level and collating them across the country. Prof Sunthar’s colleagues from JNCASR and IISc were at that time developing a model for COVID-19 growth prediction. The plan was to have an app that could not only collect ground-level data but also give local predictions in the short-term of the gap in available resources against the projected numbers.  This came out of their discussions with a district magistrate (or a collector) through the contact of his colleague. The team approached the office of the Principal Scientific Advisor (PSA) for some guidance for the project.

However, it turned out that the first part (of data collection) would be an administratively difficult task, given that everyone at the hospital and district level were so tied up with handling the initial surge of cases (this was in late March 2020). So they decided to roll out only a web-based application that would give district-wise projections.

The interface 

The website link is

Working of the application

The underlying model uses the common trends seen in available data for cases and mortality, worldwide and in India, to make predictions at the district level. Through the PSA’s office, the team was in touch with highly ranked medical professionals in the Armed Forces Medical Services. They helped in planning the inventory requirements. Using the predictions of mortality and the distribution of severity of cases (intensive, acute, and supportive), the platform is able to give numbers for the requirements of various resources (staff, equipment and consumables). Since the model uses parameters from the general trends and applies to the current cases and mortality, it becomes adaptive by nature. So the daily projections from the website are always up to four weeks from the latest available confirmed data at the end of the previous day (from The web-app, developed on AngularJS platform simply uses the raw data API  from and provides the predictions that are computed on the fly on the users’ web browser. There is no backend database or server-end computations.

On-field performance

The PSA’s office helped in showcasing this project on their webpage and twitter handle. Only a few states (AP and Punjab) have approached them directly so far. The actual data seen on the ground have thus far been within the bounds of the model projection. The web-site functions largely unattended because there is nothing much to tweak, other than API changes at the source. Currently, there is a mismatch with the live data, which may be due to API changes and the team is looking into it.

Future Plans 

According to Prof Sunthar, it was a good learning exercise for the team, working with some real data and top medical professionals.  So far it has mainly been an academic exercise. They hope to use these new contacts and the experience to make an attempt at data integration for other chronic diseases such as cardiovascular diseases and TB.  

We congratulate Prof P Sunthar and his team for this great initiative. Initiatives like this are perfect portrayals of the power of data and how data empowers us.

The content on this website is strictly the property of Insight and the Students’ Gymkhana IIT Bombay. If you wish to reproduce any content herein, please contact us:
Chief Editors: Amogh Gawaskar and Suman Mondal

Mail to: