Amazon SageMaker now supports geospatial ML (Generally Available)

Amazon SageMaker now supports geospatial machine learning (ML), making it easier for data scientists and ML engineers to build, train, and deploy ML models using geospatial data. Today, the majority of all data generated contains geospatial information, but only a small fraction of it is used for ML because accessing, processing, and visualizing the data is complex, time consuming, and expensive.

SageMaker’s new geospatial capabilities simplify the process of building, training, and deploying models with geospatial data. You can now access readily available geospatial data sources, efficiently process or enrich large-scale geospatial datasets with purpose-built operations, and accelerate model building by selecting pretrained ML models. You can then analyze and explore the generated predictions on an interactive map within SageMaker and share and collaborate on results. You can use SageMaker geospatial capabilities for a wide range of use cases, such as supporting sustainable urban development, maximizing harvest yield and food security, assessing risk and insurance claims, and forecasting retail site utilization.

Starting today, SageMaker geospatial capabilities also support Amazon Virtual Private Cloud (VPC) and AWS Key Management Service (KMS) customer managed keys. Using Amazon VPC, you have full control over your network environment and can more securely connect to your geospatial workloads on AWS. AWS KMS customer managed keys offer increased flexibility and control by using your own keys to encrypt geospatial workloads.

Amazon SageMaker support for geospatial ML is now generally available in the US West (Oregon) Region.

To learn more about geospatial ML capabilities, visit the webpage, view our documentation, or read our blog post.