Automation is proving its efficacy in exercises and pilot programs.
Pentagon officials this week highlighted the potential of emerging technologies like artificial intelligence and machine learning for automating the decision-making process at the tactical edge.
The Defense Department’s digital modernization strategy, released in 2019, pinpointed artificial intelligence and machine learning in its four main goals as essential for enabling information sharing across the enterprise. A year and change after the strategy’s release, several Pentagon officials say they are seeing real applications for edge technologies powered by AI and ML.
“The ability to take something on the very far edge where it needs natural language processing, or some forms of I say situational awareness of intent of beyond what's in front of them and have an ML trying to provide a better decision-making process on the edge of the battlespace to inform, I think those things are needed today,” Alan Hansen, chief of intelligence systems and processing at the U.S. Army’s Intelligence and Information Warfare Directorate, said. Hansen spoke Thursday at a Booz Allen Hamilton event on enabling data at the edge.
Other officials said they see some of these technologies already working in exercises. Speaking at the same Thursday webinar, Col. Charles Destefani, chief data architect at the U.S. Air Force chief data office, said he’s applied natural language processing in military exercises to capture data, text and voice streams and put them into a common field to be analyzed and used to highlight different issues and areas for the battlefield commander to act on.
The Defense Department’s recently released enterprisewide data strategy emphasizes the necessity of Joint All-Domain Operations and stated Joint All-Domain Command and Control, or JADC2, must work with the Joint AI Center and the Deputy CIO for C3 to coordinate information with the tactical edge. Destefani said JADC2 exercises are showing how automation can shorten decision cycles.
“There have been real demonstrations, I think they've had four on-ramp activities now, where they can take information from a sensor, feed it into a grid and have operational algorithms, interact with that data, present decisions to commanders, and those commanders act on those data-driven decisions in seconds, rather than minutes or hours,” Destefani said. “So, very much looking at robotic process automation and automating the decision processes at the tactical level as much as possible.”
U.S. Army Cyber Command’s Mark Mollenkopf said automation for executing pre-approved actions based on recognized conditions is a priority for the future. And in response to a question asking what AI and ML developments he found most exciting, Mollenkopf pointed to edge applications.
“I think we'll be able to put a lot of concerted effort developing models, algorithms and applications at a centralized level to be able to push that down to an edge platform where there's just not a lot of horsepower on the device, but it gives them the freedom and mobility to be able to get their military mission accomplished,” Mollenkopf said
Mollenkopf spoke during a virtual panel discussion as part of Fortinet’s 2020 Security Transformation Summit Thursday. Mollenkopf said recent pilot efforts have demonstrated the utility of automation.
Defense Undersecretary for Acquisition and Sustainment Ellen Lord struck a similar tone Tuesday in pre-recorded remarks delivered at a MITRE Corporation event launching a new 5G consortium. Lord also touched on the importance of AI and ML, particularly in environments with limited connectivity.
“This massive amount of data is a key to unlocking further technological gains in the form of artificial intelligence and machine learning,” Lord said. “Low latency communications will enable new generations of unmanned and autonomous weapon systems across all domains, empowering the warfighter with far richer access to data at the tactical edge, so that even small units can achieve strategic effects.”