JAIC develops first-of-its-kind “integration layer” for AI algorithms – Breaking Defense Breaking Defense
The Pentagon’s Joint Artificial Intelligence Center is developing a joint operating system to help combatant commands quickly build and use artificial intelligence algorithms, the agency’s director said Tuesday.
Effort is part of Artificial Intelligence and Data Initiative (AIDA) announced last year by Undersecretary of Defense Kathleen Hicks. Through AIDA, data and AI teams would be dispatched to the Army’s 11 combatant commands to help them understand the data they use and build AI tools to streamline decision-making. decision. Since announcing the initiative, the Department of Defense has completed its first round of assessments at COCOMS.
The JAIC is currently developing an “integration layer” that will facilitate the development of AI algorithms through the AIDA initiative, which has a three-year timeline. This will help COCOMs meet their “priority information requirements in their decision-making processes,” Lt. Gen. Michael Groen, director of the JAIC, told Breaking Defense following a panel at the conference. NDIA 2022 Expeditionary Warfare.
The integration layer will be the first of its kind for the Defense Ministry, he noted.
“If you want to build apps, you need a place where you can access the data feeds you need, for example, or link to other apps that you might want to integrate with your app. “Groen said. “So if you don’t have such an integration layer, you have to find all the data sources yourself. … That’s something we don’t have in the department yet. We need to be able to make app building (app development) very easy and fast by providing the backend resources and link you need to build an app.
“So it’s a pretty simple concept, but it allows us to scale because once you have all these services available, it’s very easy for anyone to develop apps and we can quickly grow our apps from AI across the force,” he continued.
RELATED: DoD Launches New Zero-Trust Security Portfolio Office; The IOC talks about priorities
The effort could assist COCOMs when creating a series of situational awareness applications or a common operating picture for a commander and additional data flow needs to be integrated. Groen said having this integration environment where the data stream is available would “just plug that into the application.”
“If you have to find that data stream and manage it separately, there’s still a lot of work to be done to make it part of your common operating picture,” he said. “So it simplifies the hard stuff and speeds up development so you can build more apps for more functions in a command headquarters, for example.”
Groen told Breaking Defense that JAIC has partnered with several vendors to develop the integration layer, but he declined to name specific companies. Other providers should be involved in the future.
“We all know the tech giants and we all know how data and AI drive finance and production, business activity, retail sales, delivery, I mean all of these companies are run by data and artificial intelligence,” says Groen. “We want the Ministry of Defense and our combatants to be able to benefit from this same level of technical integration so that they can make decisions quickly, so that they can gain efficiency on the battlefield and so that they can be productive. It’s a really powerful transformation, it just takes time to transform a ship as big as the Ministry of Defense into the tech wind so we can really start accelerating our combat capabilities.
The JAIC last month awarded Xcelerate Solutions a five-year global purchase contract for the development and integration of commercial AI tools in support of the DoD, according to a press release. Areas of focus include emerging AI technologies and include four task commands, the latest of which calls for alignment and integration with JAIC’s Joint Common Foundation, an AI development and experimentation platform. Cloud-based AI.
Groen also declined to specify the amount of funding JAIC requested for the project over the three-year period, but said the figure was in the millions.