How AI is transforming geospatial intelligence
Key takeaways
- AI reduces chart production from hours to minutes
- Human analysts remain central to quality assurance
- RAVe achieves over 99.99% first-time-right accuracy
- The technology is expanding beyond maritime mapping
When most people think about maps, they think about directions on their phone. But the U.S. Navy and other commercial vessels rely on charts and maps to show exactly what is in the vast ocean.
It’s a matter of critical safety: a navigation error at the wrong location means a collision at sea or running aground. And the volume of that maritime data is growing faster than any team can maintain.
If you can’t trust the intelligence, you’re going into a mission blind. Our data is made up of tens of thousands of different data points that are necessary to safely navigate. If the locations of any obstructions [are] off or we miss features entirely, it could spell disaster for the Navy.
Nick Cooper
Quality Assurance Manager, Leidos
In the second episode of Outcomes Unlocked, Leidos explores how artificial intelligence is changing that work.
The tool at the center of the story is RAVe — short for Raster Automation to Vector. RAVe is AI-enabled geospatial intelligence (GEOINT) software that reads complex imagery, such as nautical charts, and converts it into geo-referenced data for the maritime safety of navigation mission. Traditionally, analysts transcribed these massive datasets by hand, manually identifying and attributing tens of thousands of features and data points on a single chart. This was painstaking, repetitive work — the kind that takes a heavy cognitive toll. When you are reviewing tens of thousands of soundings, there is only so much time before something critical could be missed.
Why AI is essential for modern geospatial intelligence
This was never just a workflow problem. The world geographically is always changing, and the volume of geospatial data is vast, often unverifiable and challenging to review with the naked eye. Yet military commanders and the Intelligence Community (IC) analysts who support them need data they can trust, and they need it quickly. Solving that at a global scale is exactly where AI and automation come in.
The problem right now is [mapping and analyzing the data] takes too long, it’s too expensive, it’s monotonous, it’s repetitive in nature... the time it takes to take something process it and get it out to the warfighter is unacceptable in today’s standards.
Scot Shiflett
Director, Geospatial Innovation Center, Leidos
AI-powered geospatial intelligence results
The transformation, by the numbers:
~50% reduction in touch labor — work that was unattainable through manual production
Over “four nines” (99.99%) accuracy, first-time-right, using automation
What once took up to 40 hours now takes minutes
Leidos set out to automate the most repetitive part of that work – training AI to read charts and convert their features into validated vector data. When the team ran their first chart through the RAVe software, they were blown away – not just by the speed, but also by the accuracy.
Realizing this could be done at scale was the moment they knew it could change the way the work gets done.
How AI helps geospatial analysts work faster and smarter
Just as important as the speed is what RAVe frees analysts to do.
Instead of manually clicking through tens of thousands of features, they can focus on the higher-value analysis that drew them to the work in the first place. Analysts take pride in work that is finally focused on judgement rather than the mundane. And that human judgment is not going away. AI gives analysts a powerful tool to have in their arsenal, but the data delivered to the customer will always depend on human discernment — that part will never be replaced.
What's next for AI-powered maritime mapping and navigation
RAVe is only the beginning. The team is already extending it beyond maritime charts toward validation against real-world imagery — electro-optical and infrared data from satellites, aerial photography and point-cloud data — and toward broader geospatial applications across aeronautical and geographic missions.
The world is moving at an incredible pace, but so is our ability to understand it. And in the end, this work exists for one reason: to deliver accurate, current data that helps service members navigate safely and return home.
What excites me about this technology is it’s allowing the human to be the most efficient version of themselves to produce the best product faster than we’ve ever done before, higher quality than we’ve ever produced before and at a speed that has been unimaginable until now.
Scot Shiflett
Director, Geospatial Innovation Center, Leidos
