GIS-In-A-Box

Oct 28

Geospatial Data Infra

Tons of Earth Observation (abbr. EO) data is being acquired on a daily basis by great companies like Maxar Technologies, Planet, Umbra, among others. However, some of these companies are struggling with growth. One of the reasons is that the supply of such EO data is much larger than the demand. In fact, the demand is low because it is really hard to extract business insights from such data. Anyone who has worked with large-scale EO collections knows, pre-processing is a non-trivial lift. Much effort spent cleaning signals, harmonizing sensors, all towards the goal of producing features/inputs that are useful for downstream process models/analysis. Typically a team who is looking to extract insights from overhead EO imagery using ML would need to:

-Deploy a distributed runtime to scale out workloads such as data loading, preprocessing, and inference.

-Develop functionality to operate on raster metadata to easily filter it by location to run inference workloads on specific areas of interest.

-Optimize models to run performantly on GPUs, which can involve complex rewrites of the underlying model prediction logic.

-Create and manage data preprocessing pipelines to normalize, resize, and collate raster imagery into the correct data type and size required by the model.

-Develop the logic to run data loading, preprocessing, and model inference efficiently at scale.

Fused abstracts this complexity so you and your team can easily perform inference on massive raster datasets.

https://www.fused.io