Summary
Students at the GRILL® worked with the AFRL 711th Human Performance to develop a research tool that measures the time between auditory and visual detection. The virtual environment allows researchers to conduct controlled, repeatable studies without many of the logistical and environmental limitations of physical field testing. The application features researcher and participant interfaces that run on PC and in virtual reality (VR), respectively.
Features and Design
Software
Unreal Engine 5
Blender
GitLab
_
Hardware
HTC Vive Pro VR
Contributions
Team:
Keshav Anand
Alexander Howell
Felix Sun
Mentor:
Jon Diemunsch
Zach Galpin
_
Clients:
AFRL 711th Human Performance Wing, Human Effectiveness Directorate
Dr. Frank Mobley
Alexander Hough
Rachel Vickhouse
Outcomes and Additional Information
The program allows researchers to customize scenarios based on environment, sUAS model, flight path, weather, and participant starting position. Six quadcopters can be used across five different environments.
The application records time-synchronized data throughout each trial, including participant and aircraft positions and orientations, the participant’s viewing direction, and timestamps marking when the participant first hears the sUAS, first sees it, can no longer hear it, and can no longer see it.
Researchers can access the recorded data through logs for quantitative analysis and through a 3D replay that reconstructs the aircraft’s flight, the participant’s movements, and the participant’s viewing direction at the original speed. The 3D replay provides additional context for understanding how auditory cues guided the participant’s visual search during each trial.
By providing precise control over sUAS altitude, speed, and sound, it effectively assesses participants' identification and tracking capabilities.
Future work will include adding more sUAS models, acoustic profiles, and flight-path options.
Images from the virtual environment, including (from top left clockwise) an open area, desert, ship; and an image of a quadcopter.