Soar Technologies, Inc. (SoarTech) has been awarded a contract to support the DARPA Air Combat Evolution (ACE) program. A leader in autonomy and intelligent human-machine interfaces, SoarTech will be developing Air Combat Evolution – Trust and Reliability in User-System Teaming (ACE-TRUST) in response to ACE’s technical area two. Collins Aerospace, Raytheon Technologies, and University of Iowa Operator Performance Laboratory will be supporting SoarTech on this program.
“From SoarTech’s beginning, we’ve thrived in applying AI to hard problems within the military,” said Andrew Dallas, Vice President of SoarTech’s Autonomous Platforms. “ACE-TRUST builds on SoarTech’s twenty-plus years of expertise in adaptive, intelligent human-machine interfaces. What has made us successful in the past, and what continues to be our success today, is our AI. We are able to replicate the way humans function; we allow operators to interface with the AI in natural, human-like terms; and we provide an environment to share knowledge so that when an AI and human interact, they have a common grounding.”
As a whole, the ACE program seeks to address and increase operator trust in combat autonomy in a complex, human-machine collaboration environment coined “mosaic warfare.” Using human-machine dogfighting as its entry challenge problem, ACE aims to extend human operator capabilities by assigning the human the high-level, cognitive role of mosaic system mission commander and allowing the unmanned systems to engage the lower-level, individual tactics.
The ACE program is segmented into four technical areas (TAs) that feed into one another and scale appropriately through each phase, starting with modeling and simulation demonstrations, moving onto unmanned aerial vehicles (UAVs), and ending on a representative full-scale aircraft.
SoarTech’s main focus in ACE Technical Area 2 is on the pilot’s trust of the dogfight autonomy. Through pilot interactions with the system and from physiological data, the SoarTech team’s ACE-TRUST will measure the pilot’s trust, assess that trust against the capabilities of the autonomy, and work deliberately to calibrate the pilot’s trust appropriately to the capabilities of the dogfighting algorithms.
Distribution A: Approved for Public Release, Distribution Unlimited.