Aviators engage in a variety of outdoor activities where their health status, the environment, and the degree of workload and fatigue affect their performance. An innovative tool has been developed, which supports the real-time health monitoring of pilots using new algorithms based on intelligent clustering techniques for the recognition of possible health problems in flight. The Smart Profiler and the Intelligent Advisor modules of this system exploit the use of knowledge based expert systems and intelligent classification techniques. Coupled with the Portal, which also exploits the use of intelligent clustering techniques, it estimates the pilot’s performance in unknown environments. The new system targets recognizing possible problems at the time of flying, but it can also be used for the monitoring of the pilot performance and progress throughout a period of time, as it stores information from different flying sessions. The system was applied in 20 private pilots during the flight of a Cessna 152 aerobatic. The device was reliable and user-friendly, enabling us to monitor real-time health status of aviators in order to detect possible problems caused by the actual environmental conditions to which individuals are exposed, thus contributing to their health and safety in their working environments. Despite the automation and increasing technological complexity of modern aircrafts, the human operator still plays an important role in controlling those demanding systems. Piloting an aircraft is a highly complex task that requires the pilot to be proficient in numerous skills (Wilson & Eggemeier, 1991) in a hostile environment of cabin pressure changes and circadian rhythm disturbances particularly in long duration flights. The resulting overload of the pilots mandates the need for real time health telemonitoring (Charles, Winget, Charles, De- Roshia, Markley, & Holley, 1984; Denison, Ledwith, & Poulton, 1966; U.S. National Research Council, 2002; Ustinaviciene, Obelenis, & Ereminas, 2004). Real time health telemonitoring would be crucial to early detect and prevent conditions affecting aviator’s vital signs and cognitive performance.
Constructing a knowledge-based heterogeneous information graph for medical health status classification
