AMAZON
A discussion on a groundbreaking initiative from the National Geospatial-Intelligence Agency (NGA) — a $700 million investment in data labeling to empower artificial intelligence in geospatial intelligence.
NGA’s data labeling initiative is a pivotal part of its mission to harness the power of AI to enhance U.S. defense and intelligence operations. In this session, our panel will dive into how structured, labeled geospatial datasets—like satellite images, maps, and sensor data—are transforming intelligence and operational support.
Discussion Topics:
– Data Labeling for Machine Learning: How annotated geospatial data drives AI accuracy
– Impactful Applications: NGA’s initiative to improve object detection, environmental monitoring, and infrastructure classification
– The Future of Geospatial Intelligence: How NGA’s $700M initiative is setting new standards in defense and intelligence
Time Markers
00:07 – Welcome and introduction
01:30 – Setting the stage, what exactly does Geospatial Intelligence encompass?
05:37 – Why is the NGA $720 million investment so significant?
08:28 – JC, can you discuss the nature of the potential of AI, ML and the scope of how much data that we’re dealing?
12:34 – Let’s discuss risk analysis
16:47 – How do we move forward in a responsible manner?
19:29 – Talk about the blend of data labelling, using AI, human oversight?
21:58 – On pattern recognition, how do you safeguard the process?
23:51 – On surveillance and privacy, how do we develop models, not only ethically, but accurately?
26:22 – Let’s discuss speed, and needing to analyze the data quickly, and the issues being faced.
28:27 – JC, talk about the expanding commercial opportunities? And where this heading?
32:21 – Gary, let’s talk about speed again, it seems to keep coming up.
35:36 – Response analysis, what are the conversations driving that? Where’s this going?
37:57 – David, where are we at 5 years, 10 years out?
40:13 – JC, what do I need to know if I’m a young professional or student and I see this incredible pace in technological innovation?
43:55 – Gary, same question for you
46:26 – David, same question for you
48:40 – From the audience: Where does the assessment of future probability to enable advantageous action happen, to make truly geospatial intelligence?
52:09 – From the audience: How is NGA building systems to automatically run models on all that imagery automatically daily, not just limited area of interest?
53:47 – From the audience: With petabytes of data coming in, theres need to be a more efficient way of processing that data rather than as just raw image data. For pattern of life modeling are there more efficient ways of representing all the relevant and important data points in order to optimize and reduce computation overhead?
56:42 – From the audience: For all the geospatial intelligence capability can do regarding life and phenomena on earth, is there an equivalent “spacespatial intelligence” function that can do the same thing in the space domain? Geospatial intelligence enables the fight in maritime, land and air domains. Can “spacespatial intelligence” support the fight in the space domain?
01:00 – Anything we haven’t discussed you want to close with?
Moderator
– David Ariosto, SpaceNews Host, Journalist, and author of the upcoming Knopf book “Open Space”
Panelists
– David Gauthier, Chief Strategy Officer, GXO, Inc
– Gary Dunow, Vice President,USGIF
– John-Isaac (JC) Clark, Vice President, Head of Insights & Analytics, Maxar
Tune in to future webinars:
https://spacenews.com/webinars/
Sign up for a SpaceNews newsletter
https://spacenews.com/newsletters/
