SoarTech Awarded Contract for Automatic Explanation of Unmanned Air Systems (UAS) Behavior
Soar Technology, Inc. (SoarTech) has been awarded a contract to support an Air Force Research Lab project with the 711th Human Performance Wing’s Supervisory Control and Cognition (AFRL-RHCI) branch to develop technology to improve airman-autonomy teaming. SoarTech will develop an unmanned aerial system (UAS) autonomy and control station aid called Episodic Memory Reconstruction for UAS Behavior Explanation (EpEx), and will be supported on this effort by the University of Toledo, Embry Riddle Aeronautical University, and UAS Consulting. EpEx will combine SoarTech’s prior capabilities in adaptive human-machine interfaces, including the Weaver adaptive display manager and prior work in computational models of human episodic memory, with new algorithms for episodic memory analysis and visualization. EpEx will enable operators to better understand UAS decision-making when the UAS is operating in autonomous modes, enabling the Air Force to more easily test and field autonomous systems and to improve the range of missions over which autonomous systems can be operated.
EpEx will also integrate and correlate large volumes of internal UAS records with external observations to identify and characterize critical UAS decisions. The primary challenge for the operator is the difference between information available locally to the UAS, at the time of the decision, to information available locally to the operator, at time of an after-action review.
“SoarTech is excited to help the AFRL’s 711 HPW/RHCI further the Air Force vision of airman-autonomy teaming, developing technology that can significantly improve the control of a wide range of DoD and commercial unmanned systems. EpEx enables us to extend our breakthrough work in artificial intelligence and autonomous behavior explanation to achieve the goals of improving UAS operator awareness and of reducing the cost of UAS verification and validation,” said Andrew Dallas, SoarTech’s Vice President Federal Systems.
For more information about this award, please contact Lindsay Bayles, firstname.lastname@example.org