Abstract
The objective was to characterize visual scanning of a jet fighter cockpit by pilots with different air force experiences using the eye movement monitoring method and flight simulator under G-Force. Total, 28 pilots with various experiences in piloting MIG-29 participated in the study. The fewer experience subjects of 13 have on average 40h of air force training on the jet fighter, whereas the other 15 spend over 860h in MIG-29. Pilots performed basic manoeuvres required for patrolling the air space, including taking off, turning, landing. Experiments were conducted using an overload centrifuge for the MIG-29 flight simulator. The vision system of the simulator provides a wide field of view (120x70°) and high-resolution images required to visualize the terrain and aerial situation in all lighting and weather conditions. The visual scene was divided into 22 ROI affiliated to cockpit’s instrument and out of widow area. Eye-tracking was facilitated with The GLASSES portable google sensor manufactured by Sensomotoric Instruments GmBH (SMI, Tetlow, Germany). ANOVA of total dwell time and average fixation duration recorded in each flight phase were run to evaluate statistical significance between the expertise at 95% confidence (p<0.05). Significant differences in total dwell time between novice and experts were revealed while patrolling air space regarding attitude director indicator (ADI) and exhaust gas temperature (EGT). The median duration of fixation was also significantly different on Altimater while turning. Fixation duration on airspeed, EGT, IPV were unique for each group during approach landing. Also, the total duration time on the altimeter and ADI were significantly different. Pilots with various experiences sample visual information from the cockpit differently. Shorter fixations were not necessarily correlated with expertise, as experts dwelled longer on selected cockpit instruments during landing and turning.