The mission of the United States Air Force’s Seek Eagle Office (AFSEO) is to ensure flight safety. To fulfill this, automation and machine learning have become critical components, helping the agency to meet its mission by completing aircraft assessments more effectively and, thereby, protecting today’s warfighters.
“At AFSEO, we deliver state of the art capability to the field,” explained Donna Cotton, chief data officer, AFSEO, during the Data as a Strategic Engineering Asset Webinar “We really value being a responsible steward of our resources and it is critical to our leadership. To be effective, we must be agile, trusted, and responsive.”
According to Cotton, AFSEO receives a few hundred requests to fly new aircraft per year. From there, it’s the office’s responsibility to formally recommend whether the aircraft can be run safely and under what conditions. They must also provide documentation of the engineering rationale for the recommendations.
With all of these requests, Cotton shared that the AFSEO needed more effective ways to make recommendations and reduce the burden on the office’s workforce. This means automating manual processes where possible, such as assessing structural damage and managing aircraft storage and separation. The process of issuing a recommendation includes analyzing many complex and critical aspects of each aircraft, including flutter. Flutter, which refers to the vibration of a plane’s wings, is critical to understanding if a plane is safe to fly.
Assessing flutter is one area where AFSEO is putting automation and machine learning technologies to work. “We’re using machine learning to predict the amount of vibration in the wings as an actual number,” explained Cotton. “If you can predict this number accurately, then the decision about flight safety is so easy.”
To determine vibration, experts at AFSEO combine data from all flight simulations and enter it in a machine learning model; predictions then come out of this model, including the speed at which the plane is flying and its altitude. This data is then automated by DataRobot, which cycles through possible classifiers and chooses the best model. “The fact that we’re automating automation [to help determine safety] is amazing to me,” explained Cotton.
The mission of AFSEO is one that requires precision. In protecting today’s warfighters, AFSEO must analyze all aspects of requested aircrafts, including flutter. That’s why automation and machine learning are critical to the aircraft’s approval and recommendation process. With these advanced technologies integral to AFSEO initiatives, the process of guaranteeing flight safety is made easier.
Ready to learn more? Click here.