AFRL-Funded Researchers Conduct Detecting Deception Study

  • Published
  • By HQ AFRL
  • HQ AFRL
WRIGHT-PATTERSON AIR FORCE BASE, Ohio --  AFRL-funded scientists from the University of Arizona used video and audio tools to collect data for interpreting the body movements, voice inflections, and linguistic features of volunteer human subjects during interviews and interrogation sessions. The team used the collected data to develop a software tool suite comprising prototypes for communications analysis, deception detection, and relevant training aids.

One suite component is Agent99, which focuses on deception and its detection--addressing not only the identification of reliable indicators, but also the automation of processes that can help humans make better judgments. Continued research building upon the Agent99 concept will benefit the Air Force (AF) and other armed forces--as well as intelligence agencies--in a variety of ways. For instance, one important application lies in assisting security professionals who work in security screening situations.

The tool's namesake is "Agent 99," a character from the 1960's television series Get Smart. The prototype software suite also contains the Agent99 Trainer, an interactive mentoring/training tool that teaches law enforcers how to detect deception. This trainer was deployed as a training tool for a select group of AF officers. Agent99 evaluates vocalics [articulation], linguistics, and body language based on predefined cues that scientists have developed and programmed into the tool software. The Agent99 Trainer helps trainees understand these cues and also evaluates their level of understanding.

Researchers envision the use of this technology in a futuristic airport where passengers interact with a kiosk capable of detecting deception as their first step in the passenger screening process. While these kiosks would be similar to those currently enabling passengers to obtain boarding passes, they would have the added capacity to triangulate information from multiple databases to determine a specific passenger's risk before that person enters the next phase of security. The envisioned kiosk would covertly communicate any threat to the second security stage without alerting the passenger.