Predicting medical inventory, asymptomatics and lockdown effects

Team Lead: Aloke Kumar


Medical Inventory short-term projection with district-level granularity


This multi-institutional project involves the creation of a district-level short-term forecast of medical inventory dashboard for COVID-19. This includes inventory for intensive, acute supportive care requirements. The dashboard provides projections down to the district-level about important medical requirements such as PPEs, ventilators, oxygen and masks. Data can be used by public and administrators alike.


In the above figure, in the left hand side is shown a heuristic adaptive model which can be used for prediction of disease trajectory from time-series analysis of data. Such projections can provide reliable estimates for time-horizons of 2-6 weeks and can be used for calculating critical requirements such as medical inventory. In the right, we show a snapshot of our medical inventory projection platform, which can be accessed at

Current Status:

Project was concluded in May 2020. This project has now been superseded by a different project, which captures the epidemic through a tractable system of mathematical equations.


Team members:

Prof. Santosh Ansumali (JNCASR), Prof. Meher Prakash (JNCASR), Prof. P. Sunthar (IIT Bombay), Prof. Aloke Kumar (IISc)


Prakash, M.K., Kaushal, S., Bhattacharya, S., Chandran, A., Kumar, A. and Ansumali, S., 2020. Minimal and adaptive numerical strategy for critical resource planning in a pandemic. Physical Review E, 102(2), p.021301.


Predictive modeling of Asymptomatics, Lock-down effects and Medical Inventory

Figure 1. Active Infection prediction for Bangalore. This simulation was performed in July.

Figure 2. Active infection predictions compared with real data for Delhi. Simulation shows the impact of lockdown. Simulation was performed in July.


The multi-institutional team has developed a model, which can account for hidden asymptomatics and a predictive model for medical inventory for the nation.


As the COVID-19 pandemic unfolded in the past months, it left the public with a sense of imminent threat, the Governments across the world with unpreparedness, all of which was compounded by the scientists who were grappling with the evolving epidemic learning piece-by-piece about its epidemiology. The several facets of the disaster management included – an adaptive modelling of the epidemiology to predict the infections; solving the inverse problems of parameter estimations from the evolving data from the different regions of India and modelling the infectious disease spread; interfacing with Governmental agencies with forecasts of the infections, critical health care requirements, fatalities; planning the deployment of critical health care requirements in collaboration with medical and defence personnel; providing inputs on techno-commercial aspects of non-traditional AI based-diagnostic technologies to ICMR. Many came forward to help quantify the problem and in the process, large amounts of data were generated in the process.

The JNCASR-IISc-IITB team have developed a robust algorithm, which could be used for estimating asymptomatic patients, and predict the medical inventory required.

Current Status: Ongoing

Team members:

Prof. Santosh Ansumali (JNCASR), Prof. Meher Prakash (JNCASR), Prof. P. Sunthar (IIT Bombay), Prof. Aloke Kumar (IISc)


Kaushal, S., Rajput, A.S., Bhattacharya, S., Vidyasagar, M., Kumar, A., Prakash, M.K. and Ansumali, S., 2020. Estimating Hidden Asymptomatics, Herd Immunity Threshold and Lockdown Effects using a COVID-19 Specific Model. arXiv preprint arXiv:2006.00045.

Scroll Up