Air Force Data Hackathon Explores Generative AI Applications in Test Processes

The U.S. Air Force Test Center’s latest Data Hackathon focused on using generative artificial intelligence to enhance various test processes.

Sponsored by AFWERX, the four-day event’s goals were to identify a business use case for the technology and create a knowledge base to implement it.

The Air Force is restricted to publicly releasable data when handling generative AI. In addition, it is only allowed to prompt models such as ChatGPT using unclassified information.

According to the service, six different open-source generative AI systems were hosted in a closed environment during the Data Hackathon, enabling the use of controlled unclassified information to provide a specific flight test context.

Maj. Riley Livermore, a flight test engineer with the 413th Flight Test Squadron, explained that AI modeling can streamline test planning and reporting. He added that the capability could find use throughout the Air Force, DVIDS reported Tuesday.

The Data Hackathon saw 44 participants from the Arnold, Edwards and Eglin Air Force Bases; the Air Force Test and Evaluation Center; and the Air Mobility Command.