An approach for safer navigation under severe hurricane damage

摘要

The US are increasingly prone to sustaining severe damage from climate change as hurricanes are predicted to become more severe. During and after hurricanes, navigation around flooded areas is a primary objective for both first responders and hurricane victims. Printed or cached maps, because of their static nature, are not ideal for conveying network availability that will be constantly in flux post-disaster. And online services may not be available or may provide unreliable information. We propose a relatively cost-effective decentralized sensing and navigation system for regular grid networks that allows effective and efficient navigation around flooded areas. Using the proposed algorithm, about 59% of the cars with reachable destination were successful in reaching their destinations for the worst case scenario, that is, a dynamically spreading flood. On average, the path length taken by successful cars is 33% longer than the shortest path without flooding. The results of the experiments show that the proposed algorithm has a low computational complexity which makes it a good fit for real-time safe path finding in regular grid networks and it has the potential for extension to other types of road networks.