By Bob Gourley
DigitalGlobe is widely known as a leading provider of commercial high resolution imagery from space. Its constellation of 5 satellites has generated a 60PB archive of imagery that is approaching 5B square kilometers and is in a position to precisely document global change to meet the needs of defense, energy, agriculture, and NGO customers. Given all the data DigitalGlobe collects is unclassified it enables broad sharing across the community.
I recently had an opportunity to learn about a less well known but very exciting part of DigitalGlobe– their Insight line of business. DigitalGlobe’s Insight business (based in Herndon, VA) is focused on delivering advanced geospatial analytic solutions that help its customers save lives, resources, and time. They have a long track record of developing novel approaches to geospatial predictive modeling that have helped military, law enforcement, and corporate security customers “narrow the search space” for security related threats.
As part of its commitment to move up the value chain, DigitalGlobe is investing in ways to extract information from its imagery at scale and fuse that with other sources of geospatial data to deliver actionable intelligence for what it calls “show me there” and “show me where” questions. DigitalGlobe calls this effort Geospatial Big Data Analytics. It is making Geospatial Big Data Analytic solutions available to U.S. Government, International Governments, NGOs, and Commercial Entities both as an on-premise and subscription service.
A critical, foundational piece of this solution is its use of a Cloudera enterprise data hub.
By designing the foundation of their capabilities around a Cloudera enterprise data hub, DigitalGlobe is able to build capabilities that are highly functional while still meeting the security and governance requirements of the modern enterprise.
The result? Tools that can handle massive ingest of data from multiple sources and can serve up capabilities to analysts fast enough to let analysts iterate their questions over the data set, again and again, without having to wait for results. You can see some examples in this online video.
I was incredibly impressed with the live demonstrations of these Geospatial Big Data Analytics tools. The demo showed how DigitalGlobe could exploit its imagery, high resolution terrain data, and social media to deliver unique insights. How unique? I was absolutely blown away. I have spent a career asking hard questions of people giving me demos of advanced analytic tools and know the questions to focus on, and everything I asked to be shown to me was shown and every request for clarification on how things work was appropriately addressed. This is a solid system that works for analysts. The resulting capability can have important, virtuous impact on missions of government and the business of industry.
More on the tech involved:
- As new data comes into the enterprise it is made immediately available through a high velocity streaming ingest capability based on Storm.
- A Cloudera enterprise data hub delivers built in search capabilities that give immediate benefit to analysts.
- An enterprise data hub also provides data provenance, auditing and security, all built in.
- This solution leverages an architecture where the data is processed once, stored once, but can be used by many applications and many analysts. This negates challenges in data logistics.
- The architecture leverages custom DigitalGlobe capabilities for low latency terrain analysis in the cloud.
If you are in the DC area you can see this capability for your self. Register for the DigitalGlobe/Cloudera event 11 June 2014 by following this link: www.digitalglobe.com/geospatial_analytics