Big Spatial and Spatio-Temporal Data Analytics Systems

We are living in the era of Big Data, and Spatial and Spatio-temporal Data are not an exception. Mobile apps, cars, GPS devices, ships, airplanes, medical devices, IoT devices, etc. are generating explosive amounts of data with spatial and temporal characteristics. Social networking systems also generate and store vast amounts of geo-located information, like geo-located tweets, or captured mobile users’ locations. To manage this huge volume of spatial and spatio-temporal data we need parallel and distributed frameworks. For this reason, modeling, storing, querying and analyzing big spatial and spatio-temporal data in distributed environments is an active area for researching with many interesting challenges. In recent years a lot of spatial and spatio-temporal analytics systems have emerged. This paper provides a comparative overview of such systems based on a set of characteristics (data types, indexing, partitioning techniques, distributed processing, query Language, visualization and case-studies of applications). We will present selected systems (the most promising and/or most popular ones), considering their acceptance in the research and advanced applications communities. More specifically, we will present two systems handling spatial data only (SpatialHaddop and GeoSpark) and two systems able to handle spatio-temporal data, too (ST-Hadoop and STARK) and compare their characteristics and capabilities. Moreover, we will also present in brief other recent/emerging spatial and spatio-temporal analytics systems with interesting characteristics. The paper closes with our conclusions arising from our investigation of the rather new, though quite large world of ecosystems supporting management of big spatial and spatio-temporal data.