A scalable and composable analytics platform for distributed wide-area tracking

Abstract

Smart cities are a manifestation of the Internet of Things (IoT). The push for smart cities has led to the proliferation of video cameras in public spaces [1]. London is expected to deploy 642,000 cameras by 2020. These cameras will be used primarily for surveillance of the city for urban safety. At the same time, such smart city deployments have also seen edge and fog computing paradigms gain preference over the cloud-only computing model [2]. This is due to the network proximity of the edge and fog devices to the data sources, offering lower latency access to data and lower bandwidth requirements to push high volume data, such as video streams, to the cloud.