Leveraging the Cloud to Achieve Near Real-time Processing for Drone-Generated Data
Sayani Sarkar, Michael Wayne Totaro, Khalid Elgazzar · 2019
Low-cost drones are an emerging technological area that open the horizon for intelligent new Internet-of-Things (IoT) and a host of other applications. Cloud-based online processing of unmanned aerial vehicle (UAV) captured data is an attractive and interesting option toward real-time data processing, as it accommodates viewing and analyzing data from a variety of sources, thereby making data accessibility across devices significantly smooth and easier. This results in a more effective and seamless professional collaborations. Cloud-based solutions also make information technology (IT) management easier by automatically installing software updates, storing backups and eliminating the need for high-specification and expensive computers for data processing and storage. The ability to upload more than one just data-set at a time also provides greater flexibility, while security matters are handled using log-in credentials. In the present work, we discuss a comparative study between two data processing approaches: in one case the data captured by an UAV has been processed using an on-board low-cost single-board computer; in the other case, the captured data has been off-loaded to the cloud for further processing. Processing times between these two approaches are compared, statistical analysis applied, in order to confirm the superiority of the cloud-based processing approach. Challenges inherent with cloud-based processing have also been identified, which will be the subject of future research work.