Army Pursues Sensor-Related Artificial Intelligence Effort
The U.S. Army’s Program Executive Office Intelligence Electronic Warfare and Sensors (PEO IEW&S) will soon start an effort to bring artificial intelligence (AI) and machine learning (ML) into the sensor environment. Called Project Linchpin, the effort will help the PEO IEW&S to build an operations pipeline for creating AL/ML solutions for sensing and data science, lightening soldiers sensing, targeting and intelligence surveillance and reconnaissance environments.
The organization is looking to put in place an “end-to-end development and deployment environment for AI/ML capabilities leveraging military intelligence data and supporting sensor requirements,” within the PEO’s existing technology portfolio, reported Laurence Mixon, defense intelligence senior leader at PEO IEW&S.
“We're looking at a way to have a ... government-owned environment with the data where we can develop, reuse algorithms and then push them out,” Mixon stated, presenting November 17 at the AFCEA Alamo annual ACE conference, held November 14-17 in San Antonio.
The Army’s second-in-charge for acquisitions, Young Bang, is encouraging the service to leverage more AI and ML applications, Mixon shared. Bang, a West Point graduate and former Army captain, spent several years in industry pursuing emerging technologies—AI, ML, quantum, augmented reality and other capabilities—with Atlas Research, KPMG and in the telecommunications industry—before returning to the Army.
“The principal deputy assistant secretary for Army acquisitions, Mr. Bang, came out of industry, and he's now the number two within the Army acquisition enterprise,” Mixon noted. “He is really pushing the PEO and the Army to change with respect to AI and ML.”
The project emerged from conversations and visits with Bang and the IEW&S Program Executive Officer Mark Kitz and other PEO IEW&S officials. “It came out of some conversations and visits that he had at Aberdeen with our folks,” Mixon continued. “We had an initiative with MITRE, where we had an AI/ML application, and through our conversations, we kind of gave birth to this new project, Project Linchpin.”
The PEO IEW&S is currently scoping the requirements for Project Linchpin.
As for considerations in developing AI/ML tools for sensing, situational awareness or intelligence surveillance and reconnaissance environments, Mixon said, “Training the algorithms is a challenge and getting the data. And we're hoping to decouple it from the software so we can reuse it. That was kind of the epiphany that we had with Mr. Bang a few months ago. That's where we want to go.”
Notably, in addition to Project Linchpin, the PEO IEW&S is seeking other industry solutions for its modernization efforts regarding integrated sensor architecture, the Improved Threat Detection System (ITDS), the High Efficiency Radio frequency Monitoring and Exploitation System (HERMES), and support of multidomain task forces and theater operations.
“So, in FY23, where are we going? What are we focused on?” Mixon continued. “With integrated sensor architecture, we see that as something that must be done. We have all these sensors out there, but they're not always supporting one another.”
So, in FY23, where are we going? What are we focusing on? With integrated sensor architecture, we see that as something that must be done. We have all these sensors out there, but they're not always supporting one another.
With the integrated sensor architecture, the Army wants to be able to “dynamically locate sensors” and then access sensor data from a variety of locations to improve situational awareness. In addition, soldiers would like to be able to have enabled sensors within an area of operation on the network that “talk” without the need to be physically integrated, Mixon clarified.
“We think that if we can get this architecture developed and designed, without having to physically integrate, we will be able to have significant gains in our ability to protect systems and locate threats,” he stated.
For the ITDS, the PEO is seeking advanced warnings for helicopters and other aircraft. “[We are looking] for modern, state-of-the-art threat detection capabilities for rotary and fixed-wing aircraft against air defense threat system capabilities,” the senor leader told the industry attendees. “It’s enabling us to categorize on the fly new threats that are coming and just more effectively use the existing sensors capabilities that we have.”
The HERMES modernization, meanwhile, includes harnessing signals intelligence, or SIGINT sensors for the aerial layer.
Mixon expects the PEO to put out several indefinite delivery/indefinite quantity, or IDIQ contract bids and one GSA contract vehicle opportunities in the first, second and third quarter of FY23.
With any industry solution, the PEO IEW&S is looking for capabilities that leverage open-source architectures—such as CMOSS or others—and naturally need technologies with favorable size, weight and power considerations, Mixon said.