Search for: All records

Infrastructures such as roadways, power lines, and communications networks play a critical role in our society. However, they are also susceptible to failures, especially after natural events, easily affecting large geographical areas. Predicting where and when these failures will occur with high confidence is very difficult due to the stochastic nature of such events. Nevertheless, it is possible to know which areas are more vulnerable in advance and plan accordingly. This paper aims to use just remote sensing techniques based on satellite images to detect roadways vulnerabilities to hurricanes. The framework exhibits a modular architecture that enables detecting and mapping in 3D vegetation and detecting buildings. We propose a risk function based on the information retrieved from the satellite image which can be used to create a risk map of the area. The study area has been selected in Tallahassee, Florida where a high-resolution satellite image has been acquired in September 2018, before Hurricane Michael main hit. The findings of this work can help the management teams and city responders to identify the most vulnerable regions which are under the risk of disruption and to organize the resources prior to the event. The advantages of our approach are that the entire framework can be use as an end-to-end standalone solution for risk analysis at city level and can be easily expanded with other source of data.