作者: Geoffrey Pettet , Ayan Mukhopadhyay , Sayyed Mohsen Vazirizade , Matthew Berger , Mykel Kochenderfer
DOI:
关键词:
摘要: Emergency response is one of the most pressing problems faced by communities across the globe. It is also one of the most important engagements of governments. In the last fifty years, developing statistical, analytical and algorithmic approaches to design emergency response management (ERM) systems has garnered a lot of attention. An ERM pipeline is an intricate combination of several modular sub-components that need to be individually understood in order to make the overall system work. We have identified several key challenges and worked on different components of the ERM pipeline. A major hurdle for widespread adoption of state-ofthe-art ERM techniques is the lack of an integrated, ready-to-use ERM toolset that allows policy makers to directly compare various techniques and visualize their impact. We are developing such an open-source toolset, the first of its kind, which integrates four major components of ERM–incident forecasting, strategic (long term) resource allocation, dynamic resource allocation, and dispatch–in a ready to use package. In this paper, we highlight our work on alleviating some of the major challenges of that ERM faces today, and talk about how such approaches are integrated in our tool.