This infrastructure project responds to the urgent need of managing and analyzing big spatial and spatio-temporal data. Such data is continuously produced from various devices including smart phones, space telescopes, and medical devices. The project goes beyond the idea of supporting spatial and spatio-temporal data using general purpose big data engines, by developing specialized spatial and spatio-temporal big data systems and algorithms. The hardware infrastructure acquired in this project, including new architectures such as Graphical Processing Units (GPU) and Solid State Drives (SSD), is used to study and develop spatial and spatio-temporal big data systems and techniques for various software and hardware platforms.
The software platforms studied range from traditional parallel database systems to state-of-the-art distributed computing platforms. Hardware platforms include large clusters of distributed computed nodes of commodity machines, GPUs, and SSDs. Spatial and spatio-temporal big data techniques investigated include indexing, querying, and map rendering. The project encapsulates its developed techniques in publicly available open-source free software. In addition to use within the University of Minnesota, the acquired hardware infrastructure is made accessible to developers, practitioners, and researchers worldwide through a wide set of publicly available web services for spatial and spatio-temporal big data.