AIRWAVE: Artificial Intelligence for Remote Wireless Assessment of Vulnerable Electronics

The US Army, as well as other service organizations, have a national Defense-related mission need in the area of vulnerability assessment and Counterintelligence (CI) mitigations for Department of Defense (DoD) interests. Specifically, establishing and maintaining confidence that the security infrastructure is not subject to any unknown weaknesses / vulnerabilities or foreign interference is absolutely fundamental for US Forces and our allies to maintain a dominant posture over near-peer adversaries and foreign actors. A solution to objectively assess the vulnerabilities of DOD supply chain facilities and where the DOD has interests in commercial / industrial supply chains will help to provide the assurance that is desperately needed as foreign competitors are actively posturing to inject themselves into our developmental pathways wherever possible to create the potential to disrupt our information, logistic, and technology advantages. These assessments will inform investments in countermeasures to detect or mitigate attack vectors targeting identified vulnerabilities resulting in comprehensive risk reduction. To begin to address the above problem, Knowmadics Inc. (ōKMIö) is developing multiple COTS+ RF signal detection and characterization solutions. These COTS+ sensors, as well as other GOTS and third-party COTS RF sensors are applied by EW Forces to conductĀdiscreet, repeatable, and robust network characterization and MDVA, often of a facilityÆs external surveillance and security measures or individualsÆ RF signatures (e.g., Wi-Fi, Cellular, Bluetooth on a carried smartphone). However, due to the volume of data collected even during a relatively short collection period (e.g., 8-hours), additional capabilities are needed that can make effective use of collected data in a timely manner. To overcome this problem, KMI executed a successful Phase I effort to design, develop, and conduct initial feasibility evaluation of a proof-of-concept Artificial Intelligence for Remote Wireless Assessment of Vulnerable Electronics (AIRWAVE). The objective of a full-scope AIRWAVE solution in Phase II is to provide trained and validated machine learning (ML) algorithms that ingest the spectrum of RF data collected from any third party source (initially piloted with our own COTS+ sensors, but adopting industry standard file formats to remain open to commercial, or GOTS RF collection tools) and can be used to rapidly and accurately identify risks and/or anomalous events to/for an individual supply chain facility. AIRWAVE ML models will continuously learn and adapt to a specific installationÆs security or area of interest (AOI) and RF patterns of life (PoL) to provide an adaptive solution that helps to identify potential risks earlier, and inform effective mitigations more consistently to promote supply chain resilience as part of a layered defense security plan.