Braxton Taylor
Next year, two U.S. warships are slated to get a prototype AI-powered system designed to filter their sensor data for potential targets—and then predict their behavior, according to the system’s manufacturer.
“The sensors that the Navy operates are so sophisticated and sensitive [they] can pick up very small things, but they need to be able to tell, ‘Is that a small thing that we should care about or is that a small thing that is just irrelevant’?” said Ben FitzGerald, CEO for Rebellion Defense.
Last year, the Navy’s Program Executive Office for Integrated Warfare Systems X—the outfit responsible for upgrading combat platforms—awarded Rebellion Defense a contract to continue developing its Iris target-processing software. Awarded via other transaction authority, the contract covered two prototype phases and a production phase.
Now, the Navy is extending Rebellion’s contract for an undisclosed amount and another 14 months, aiming to put Iris on two ships and ensure that it “works in a way that the Navy is comfortable with on Navy platforms and systems,” FitzGerald said. “We’ll be doing integration work through the IWS software pipeline, and then we’ll be able to do testing on ships in 2026.”
A Navy official said the service is pursuing testing plans.
The company is soliciting input from sailors as it tailors Iris to work with the Aegis Combat System and Ship Self-Defense System, he said.
Iris sits on top of existing infrastructure and helps sailors spot, track, predict, and assign weapons to targets. For example, the software can help determine whether a flying object is more likely an airliner or an armed enemy drone.
“If there is something that is smaller than [a passenger jet] and moving very, very slowly, but says that it’s a commercial airliner, you should be watching out for that,” FitzGerald said.
The same thing applies to ships.
“We want to identify, for humans, any ship that is traveling at 95% or greater of its top speed. And then you can start getting into things like,’Well, it says that it’s an oil tanker, but it’s going faster than any oil tanker has ever been known to travel. Humans should be looking at that,” he said. “Because…people are so focused on looking at all sorts of other things that they don’t see the anomalous thing.”
The move is part of a larger Navy effort to upgrade its software and improve maritime operations with access to data analytics.
Brent Sadler, a senior research fellow specializing in naval warfare with the Heritage Foundation, said predictive AI and big-data techniques can help trawl through a lot of targets, like a group of commercial fishing vessels, trying to pick up which ones might be bad actors.
“So rather than the 200 that are out there that you have to worry about, you can then focus on these. This is what they’re trying to get at in peacetime…to better assign limited platforms and sensors using AI and big data analytics,” Sadler said.
Things become more complicated and crucial in combat.
“In wartime, it becomes a lot more about target queuing, and that is: Do I have the right target? Where’s the target most likely to head? And then launch an attack on them in a way that they can’t detect until it’s too late?” he said. “It’s a little bit more urgent and time-critical for putting a weapon on a target to try to anticipate how it moves and where it might be going, but also to make sure you have the right one.”
Read the full article here