00:00 AM

Visiting Scholar Sanjeev Goyal Wraps up Yearlong Collaboration with Pitt and American Red Cross

Sanjeev GoyalPITTSBURGH (October 25, 2017) … The University of Pittsburgh hosted Sanjeev Goyal, assistant professor at the YMCA University of Science and Technology in Faridabad, India, as a postdoctoral scholar focusing on the use of predictive models in disaster response. Dr. Goyal has been working under the supervision of Louis Luangkesorn, assistant professor of industrial engineering at Pitt’s Swanson School of Engineering, to predict the demands for food and shelter services following major floods.

“In the early days of a disaster, the deployment of state and national resources into the disaster area is often delayed pending a request for resources from local agencies,” said Dr. Goyal. “The delay can be lengthened because local personnel are conducting first response operations. However, it may be possible to initiate movement of resources into an affected region if an estimate of the needs can be made, then direct specific resources as local agencies make specific requests when preliminary assessments are completed. Predictive models promise to provide these types of estimates.”

To develop a predictive model, Dr. Goyal and Dr. Luangkesorn are using demographic, physical, and historical data that is readily available outside of the disaster area on the first day of a disaster. Demographic data is represented by the Social Vulnerability Index maintained by the U.S. Centers for Disease Control and Prevention. Physical damage impacts are represented by the National Weather Service Advanced Hydrologic Prediction Service historical and current flood gauge data. Historical data is available from American Red Cross damage assessment and feeding and sheltering operations from past major floods. 

The model seeks to predict the damage to residences and the resulting needs for food and shelter. This prediction can then be used to initiate the assignment of supplies, personnel, logistics, and financial resources to disaster relief. Efforts can be refined as more information is available. Initial results of the model have been used to inform response to floods in summer 2017 in the United States. Further work is intended to prepare the model for use by disaster response agencies in making initial resource requirement estimates in areas impacted by flooding rivers. 

Dr. Goyal received support from a University Grants Commission of India for the one-year fellowship, which began October 2016. The work was done with the advice and assistance of Mr. Michael Whitehead, government operations manager at American Red Cross.


Author: Louis Luangkesorn, Assistant Professor of Industrial Engineering