Forward-looking cities across the globe must be robust against natural disasters. Such disasters demand immediate response from decision-makers to minimize the damages. Applications, such as evacuation traffic modeling, that provide real-time analytical information in emergency situation generally process large volume of data (i.e., Big Data) from sensors and cameras. Processing of such Big Data applications are often resource-intensive.
Private Cloud infrastructures play a critical role in providing resources for disaster management applications. However, such Clouds are frequently overwhelmed by the load of disaster management applications. In addition, it would be cost-prohibitive for cities to over-provision infrastructure to respond to rare disaster events. Several organizations within a community, on the other hand, have plenty of underutilized resources, and are willing to share them in disastrous situation. Besides, existence of a high-bandwidth and robust network within a local jurisdiction makes them an ideal environment for Big Data disaster management applications. Our objective in this proposal is to create a dynamic community Cloud that provides the infrastructure for Big Data disaster management applications. Such community Cloud can scale in response to increasing demands of the relief operations and then shrink back in times of normal operation. This solution deals with the challenges of ensuring response time, scalability, transparency, performance, and security. Being motivated by this scenario, this project aims to develop techniques and tools for efficient use of community Clouds to process real-time Big Data disaster management applications and make robust smart cities.
The project is funded by the Board of Regent Louisiana.
The project will officially start from August 2016.