SYDNEY J. FREEDBERG JR.
WASHINGTON: This morning, Deputy Defense Secretary Kathleen Hicks announced a new Artificial Intelligence & Data Acceleration (AIDA) initiative to help the military’s 11 inter-service Combatant Commands.
The COCOMs command operational forces around the world, so their involvement is central to the military’s plans for a Joint All-Domain Command & Control (JADC2) meta-network linking forces across land, air, sea, space, and cyberspace. AIDA will be a centrally directed DoD effort to help the COCOMs, contrasting with the disparate service-led initiatives on AI and data.
AIDA’s objective is to help the COCOMs understand what data they use to make decisions, automate the data flow and create AI tools to streamline decision-making. The plan is to make use of already-scheduled COCOM experiments and exercises to try new capabilities in real-world environments; whatever works best will be left in place permanently for the command to use.
“Through successive experiments, we seek to understand the obstacles and challenges that impair our current ability to rapidly scale AI across the department and the joint force,” Hicks told the Joint AI Center conference this morning. As we complete these episodic exercises and experiments, we intend to leave behind capability in our wake.”
The AIDA initiative will dispatch two different kinds of teams to COCOM HQs, Hicks said, one focused on data, the other on AI:
“Operational data teams” will go to COCOMs for longer-duration efforts on data. Their job is to “catalog, manage, and automate data feeds that inform decision-making” until those data feeds are “captured, completed, curated and usable.” In other words, they’ll figure out what data COCOMs are using, vet that data to make sure it’s both comprehensive and valid, and get it to move automatically to commanders and staffs in easily intelligible and readily usable forms.
Meanwhile, “flyaway teams of technical experts” will drop in on COCOMs to work on artificial intelligence tools. They’ll bring “top tier talent and technology” to apply AI to “streamline and automate workflows.” In other words, they’ll look at how the COCOMs use data to make decisions, then apply AI to improve those processes and automate what is currently laborious manual staff work.
The teams will also gather information on what kind of network infrastructure upgrades the COCOMs need and what policies need to be changed to ease data sharing. Ultimately, Hicks said, the goal is to enable free flow of data across different COCOMs using open data standards, and real-time integration of data from different sensors and unmanned systems.
No comments:
Post a Comment