Routing Algorithms as Tools for Integrating Social Distancing with Emergency Evacuation

Abstract

We explore the implications of integrating social distancing with emergency evacuation, as would be expected when a hurricane approaches a city during the COVID-19 pandemic. Specifically, we compare DNN (Deep Neural Network)-based and non-DNN methods for generating evacuation strategies that minimize evacuation time while allowing for social distancing in emergency vehicles. A central question is whether a DNN-based method provides sufficient extra routing efficiency to accommodate increased social distancing in a time-constrained evacuation operation. We describe the problem as a Capacitated Vehicle Routing Problem and solve it using a non-DNN solution (Sweep Algorithm) and a DNN-based solution (Deep Reinforcement Learning). The DNN-based solution can provide decision-makers with more efficient routing than the typical non-DNN routing solution. However, it does not come close to compensating for the extra time required for social distancing, and its advantage disappears as the emergency vehicle capacity approaches the number of people per household.

A Multipopulation Multiobjective Ant Colony System Considering Travel and Prevention Costs for Vehicle Routing in COVID-19-Like Epidemics

“…Given that the quarantine can prevent the virus spread, Mook et al [43] analyzed the efficiency of quarantine policy for VRP in epidemics. In addition, some studies also cover the issue of social distancing (e.g., the maximum number of customers assigned to each vehicle) [44],…”

Improving debris flow evacuation alerts in Taiwan using machine learning

“…Data-driven machine-learning models are increasingly employed in improving early warning models, weather and natural hazard forecasts, and disaster evacuation management. Examples include weather forecasting 20,21 , landslide displacement prediction 22 , spatial mapping of debris flow susceptibility [23][24][25][26] , predicting scales of landslides 27 and monthly rainfall for early warning of landslide occurrence 28 , differentiating between ground vibrations generated by debris flows and other seismic signals 29 , and enhancing disaster response and emergency evacuation planning [30][31][32][33] . However, to the best of our knowledge, none of the existing studies predict the occurrences of debris flows within a selected time using machine learning algorithms trained on historical hourly rainfall data alone.…”