Multi-criteria performance analysis based on Physics of Decision - Application to COVID-19 and future pandemics

The purpose of this study is to present a novel perspective on decision support based on the conventional SEIR pandemic model paradigm considering the risks and opportunities as physical forces deviating the expected performance trajectory of a system. The impact of a pandemic is measured by the deviation of the social system’s performance trajectory within the geometrical framework of its Key Performance Indicators (KPIs). According to the overall premise of utilizing Ordinary Differential Equations to simulate epidemics, the deviations are connected to several alternative interventions. The model is essentially built on two sets of parameters: (i) social system parameters and (ii) pandemic parameters. The ultimate objective is to propose a multi-criteria performance framework to control pandemics that includes a combination of timely measures. On the one hand, the current study optimizes prospective strategies to manage the potential future pandemic, while on the other hand, it explores the COVID-19 epidemic in the state of Georgia (USA). IEEE

Data-driven prediction of demand in a pandemic for critical consumables over a shared distribution network

This paper addresses a novel demand prediction approach based on a dynamic and adaptive data-driven model for critical consumables in a pandemic considering a shared distribution network. We have developed this approach as a key domain in the personal protective equipment (PPE) supply chain project for Georgia Tech (GT) research laboratories for robust inventory management and reassurance of GT's continued opening. Then we have applied this method to different group levels of users at GT (research laboratories, buildings, and the entire campus). Using agreement and collaboration for shared inventory management on campus, predicting PPE demand at different group levels is imperative since we can lower safety stock levels while maintaining the same service level by leveraging pooled inventory. During COVID-19 at GT, hundreds of research laboratories have routinely reported their PPE consumption for months to contribute to a pan-campus initiative to secure persistent supply and distribution of PPEs. Leveraging the daily information gathered from multiple sources, we model the current and projected consumption rates per person per day at each level over the forthcoming months. Then we generate simulation-based multiple scenarios of predicted consumption enabling the model to develop time-phased stochastic demand distributions. Our approach has been implemented to support PPE supply and distribution across GT campus and most of its daily updating has been automated. In the paper, we describe the aim and context, formally present our approach and model, provide empirical results, draw conclusions and propose avenues for further research. © 2021 IISE Annual Conference and Expo 2021. All rights reserved.