An AFIT-led team celebrates winning first place in the Hackers Choice Awards at the
Bravo 11 Bits2Effects Hackathon in Hawaii. (Contributed photo)
A team led by Air Force Institute of Technology faculty members Lt. Col. James Dean and MSgt. Benjamin Johnis, along with master’s student Lt. Josiah Stearns, delivered a fully functional prototype that automatically identifies malicious navigation signal interference with 96% accuracy as part of the Bravo 11 Bits2Effects Hackathon in February. As a result, the AFIT-led team received first place out of 72 teams in the Hackers Choice Awards.
“I’m exceptionally proud of the AFIT team’s performance in solving a real operational challenge under intense pressure,” said Lt. Col. Wayne Henry, director of the AFIT Center for Cyberspace Research. “Their innovative work showcases their exceptional talent and underscores AFIT’s leading role advancing warfighting capabilities.”
Hosted by the Office of the Secretary of Defense Chief Digital and Artificial Intelligence Office and US Indo-Pacific Command, the event welcomed nearly 700 participants to the Department of Defense AI Battle Labs at Scofield Barracks, Oahu, Hawaii. This year marked the fourth BRAVO hackathon and the first to be held inside a combatant command.
Hackathon Background
Starting in 2021, the Air Force began organizing multi-service prototyping events, known as BRAVO hackathons, to expedite learning and capability development from classified and protected operational data. The goal of the hackathon is to produce solutions to combatant command challenges utilizing Indo-Pacific operational theater data.
BRAVO utilizes a permissive software development environment that permits the co-mingling of classified and protected data with untrusted open-source and commercial software otherwise not approved for production systems within minutes. Prior hackathons have produced prototypes influencing major Defense Department programs in areas including large language models, space launch, flight telemetry and biometrics, unmanned systems, personnel recovery, security classification, sensing and targeting, and battle damage assessment among others.
“BRAVO Hackathons represent an opportunity for DoD to practice and proliferate the fundamentals of user-centered design and agile software development," said Joe Larson, the Defense Department's deputy chief digital and AI officer for algorithmic warfare.
The Challenge
Johnis, an instructor of operational sciences, came to AFIT from the combat search and rescue career field as a specialist in personnel recovery. One of the challenges he experienced in combat was adversary interference with GPS navigation signals. The military has equipment that allows operators to identify and characterize signals that interfere with navigation signals, but this equipment requires specially trained operators, which limits the number of sensors that can be deployed to the field. To solve this problem, Johnis sponsored the Navigation Warfare use case, challenging interested hackers to develop a machine learning algorithm that could automatically identify malicious interference and notify a central command authority.
The Solution
The resulting NAVWAR team, led by AFIT, was made up of 17 hackers with backgrounds from aircraft maintenance and communications technicians to commercial web developers.
To accomplish its goals in the short amount of time given, the team split into three sub-teams. Stearns led the team developing the back-end software that would tie the sensor, AI algorithm and reporting database together, Dean led a team in developing and training the AI algorithm and Johnis led the team developing the operator interface and coordinated with operational sponsors to ensure the prototype met the requirements for transition to operations.
At the end of the hackathon, the team delivered a fully functional prototype that was able to demonstrate a concept for the operator interface, an AI algorithm that was able to identify malicious interference with 96% accuracy, and a server application that was able to host the system in real-time. This prototype will be utilized to fast-track authority to operate on government systems and ensure the capability can be rapidly transitioned to operations.
On the final day of the Hackathon, subject matter experts evaluated each of the use cases based on their innovative use of data, the impact on the warfighter and the technical challenge, culminating with a vote to determine top-tier competitors for the Hackers Choice award. The AFIT-led team was awarded first place out of 72 teams in the Hackers Choice Awards.
In addition to delivering a functional prototype, the members of the NAVWAR team returned to their units with hands-on experience using tools and techniques for modeling malicious GPS interference, in addition to developing and training machine learning algorithms.
Information for this article was obtained in part from www.defense.gov.
