—Natural or man-made disasters wreak havoc, whether they are floods, earthquakes, tornados, hurricanes, or wild fires. One of the major challenges in emergency situations is to guide people through safer routes away from the disaster site on the basis of the available information. Data from large number of sources, such as road side sensory units, emergency vehicles, satellite imagery must be processed at real-time to compute the appropriate routes for evacuees. However, to process the huge volumes of sensory data at real-time also requires higher computational resources. For many years, cloud computing has been well-established as a most reliable solution to meet higher data and computational demands. In this magazine article, we discuss the architecture of a scalable cloud-based emergency evacuation service, namely EvacSys. The service utilizes the power of cloud-computing to process large volumes of real-time sensory data gathered during disaster and computes appropriate routes fo...