Using AI for improving evacuation during compound disasters
Abstract/Contents
- Abstract
- Compound disasters have been leading to extensive casualties and property losses around the world as more frequent and intense natural disasters, pandemics, and humanitarian crises have overlapped in recent years. Evacuating people in a more timely, efficient, and safe manner has thus become a complex challenge in disaster response and management. However, very little work has utilized AI to improve evacuation during compound disasters. The work presented here aimed to address this gap. In its three central chapters, this dissertation addresses three aspects of disaster evacuation amidst a changing climate and an emerging pandemic: early warning decision-making prior to an initiation of disaster evacuation, emergency vehicle routing during a disaster evacuation, and pandemic risk mitigation after initiation of a disaster evacuation. After an introductory chapter, in my second chapter, I present data-driven machine learning models that improve upon the existing rainfall-based debris flow warning model in Taiwan. In addition, I identify the rainfall trajectories most strongly related to debris flow occurrences and provide decision-makers with comprehensive trade-offs and multi-objective analysis to balance the risks of missing debris flows versus frequent false alerts. In my third chapter, I discuss whether a DNN (Deep Neural Network)-based solution (Deep Reinforcement Learning) can provide sufficient extra routing efficiency to accommodate increased social distancing in a time-constrained evacuation operation compared to a non-DNN solution (Sweep Algorithm). My fourth chapter shows how I used an age-structured epidemiological model, known as the Susceptible-Exposed-Infectious-Recovered (SEIR) model, to analyze to what extent different disaster evacuation strategies can decrease infections and delay pandemic peak occurrences. In sum, the data-driven findings presented in this dissertation demonstrate the utility of AI and data science as decision-support tools for disaster resilience, and contribute to model-based decision-making for the dynamic compound disaster response and evacuation operations.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2022; ©2022 |
| Publication date | 2022; 2022 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Tsai, Yi Lin |
|---|---|
| Degree supervisor | Field, Christopher B |
| Degree supervisor | Kitanidis, P. K. (Peter K.) |
| Thesis advisor | Field, Christopher B |
| Thesis advisor | Kitanidis, P. K. (Peter K.) |
| Thesis advisor | Fletcher, Sarah |
| Degree committee member | Fletcher, Sarah |
| Associated with | Stanford University, Civil & Environmental Engineering Department |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Yi Lin Tsai. |
|---|---|
| Note | Submitted to the Civil & Environmental Engineering Department. |
| Thesis | Thesis Ph.D. Stanford University 2022. |
| Location | https://purl.stanford.edu/bn052yt5523 |
Access conditions
- Copyright
- © 2022 by Yi Lin Tsai
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...