Screener Vigilance Detection

The Challenge

The job of performing airport security screening can be a monotonous visual search task characterized by long periods of watching images of checked and carry-on baggage. Fortunately, most baggage will not contain any threats for security; however, the low probability of a screener encountering a threat may make it more likely that he or she will not detect this event if it ever does occur. Screeners can therefore suffer from vigilance and cognitive fatigue problems.

Our Approach

Shorter shifts and well developed work-rest cycles can mitigate vigilance problems but are not always implemented or enforced due to mission requirements or staffing problems. Alerting the operator and/or supervisors when performance is degraded because of cognitive fatigue and vigilance lapses is an alternative solution, assuming that a feasible monitoring system can be developed and deployed. Such a monitoring system must have the following characteristics:

  1. It must be a non-intrusive system that will not distract the screener or compromise the mission of threat detection.
  2. It must operate in real-time and maintain accuracy in detecting lowered levels of screener alertness.
  3. The system performance should not be influenced or significantly degraded by environmental conditions such as differential lighting and noise levels.
  4. The system must be simple to use or wear, and it should be easy to calibrate or require no calibration prior to use.
  5. The system should be relatively inexpensive to purchase, maintain, and operate.

Recognizing these constraints imposed by the environment in which such a system must operate, we proposed a goal for development that ultimately produces a system requiring no wearable sensors. Specifically, we are aiming for a system that, when used in an operational environment, requires only an off-body eye-tracking system but one which is fine tuned using individualized indications of fatigue and vigilance derived from other physiologic measures acquired during training. To achieve this goal, we concentrated on identifying physiological measures as being the most robust indicators of fatigue and vigilance by conducting a pilot study to verify the sensors and measures used in previous vigilance studies from the literature and, finally, to integrate these sensors using the sensor fusion techniques provided by pattern classification algorithms.

Project Details

  • Sponsor: Department of Homeland Security (DHS)
  • Application: improve vigilance and combat cognitive fatigue
  • Target markets: baggage screeners, commercial truck drivers, building security, and automated assembly line monitoring