Big Data Processing

The rapid growth of both imagery and other geospatial data has increased the complexity of creating responsive and scalable geospatial applications.  Azavea is a leading expert in making high performance geospatial processing possible on the web.

Raster data sets can easily have millions or billions of values, and performing operations on them is often a very computationally expensive task. We believe that through performance-minded code and techniques such as parallelization, raster data can become a critical piece of high performance applications.

To this end, we designed GeoTrellis to support very fast geospatial data processing.  It is based on the Scala language and the Spark framework.  GeoTrellis provides the ability to process large and small raster and network data sets with low latency by distributing the computation across clusters of multiple threads, cores, CPUs and machines.  Thanks to GeoTrellis you can now solve three core challenges:

  • Create fast, scalable geoprocessing web services
  • Create batch geoprocessing services that can act on large data sets
  • Run geoprocessing operations in parallel to take full advantage of cloud computing architectures

The software includes the ability to both rapidly process geospatial data and chain together multiple processing tasks to support simulation and modeling. It literally transforms user interaction with geospatial data by bringing the power of geospatial analysis to real-time, interactive web and mobile applications.