Investigating the relationship between eye movements and situation awareness in weather forecasting

Abstract

Physiological indicators, including eye tracking measures, may provide insight into human decision making and cognition in many domains, including weather forecasting. Situation awareness (SA), a critical component of forecast decision making, is commonly conceptualized as the degree to which information is perceived, understood, and projected into a future context. Drawing upon recent applications of eye tracking in the study of forecaster decision making, we investigate the relationship among eye movement measures, automation, and SA assessed through a freeze probe assessment method. In addition, we explore the relationship between an automated forecasting decision aid use and information seeking behavior.In this study, a sample of professional weather forecasters completed a series of tasks, informed by a set of forecasting decision aids, and with variable access to an experimental automated tool, while an eye tracking system captured data related to eye movements and information usage. At the end of each forecasting task, participants responded to a set of questions related to the environmental situation in the framework of a surveybased assessment technique in order to assess their level of situation awareness. Regression analysis revealed a moderate relationship between the SA measure and eye tracking metrics, supporting the hypothesis that eye tracking may have utility in assessing SA. The results support the use of eye tracking in the assessment of specific and measurable attributes of the decision-making process in weather forecasting. The findings are discussed in light of potential benefits that eye tracking could bring to human performance assessment as well as decisionmaking research in the forecasting domain.

HMD-Based VR Tool for Traffic Psychological Examination: Conceptualization and Design Proposition

“…The device is equipped with several sensors: SteamVR Tracking, G-sensor, gyroscope, proximity, IPD sensor and eye tracking. Eye fixation on relevant objects was found to be related to driver's situation awareness [70,71], as discussed above. ET provides binocular 120 Hz gaze data output with accuracy 0.5 • -1.1 • and 5-point calibration.…”

“…Eye-tracking technology can help capture important measures, which are barely measurable by standard TPE [50]. The human sensorimotor activity including pupillometry, blink rates or fixations on specific objects in the environment (areas of interest) were within the ranges found in previous studies, that reliably indicate different states of a driver's arousal, fatigue, deception, attention change or loss, hazard perception or complete loss of control [68][69][70][71]. The eye-tracking technology was also proved to have the capacity to predict neurocognitive illnesses [72], which might speed up the decline of driving skills and hence it may promote prevention in road safety.…”

“…Based on sensorimotor activity, several cognitive and emotional phenomena can be induced. Eye fixation on relevant objects in the visual field was found to be related to driver's situation awareness [69,70]. The way drivers look differs according to their experience-experienced drivers fixate more on higher positions and for a shorter time than the more inexperienced drivers [73].…”

“…For the measurement of SA in driving assessment, we proposed the path diagram of the conceptual model including key factors relevant for TPE (see Figure 3) along with corresponding measures (Table 3). The diagram is based on the previous studies on SA in combination with ET measurement options for driving assessment [52,[68][69][70][71]. With respect to measurement limits and options of the used VR technology, the model was adjusted by including and excluding specific variables.…”

“…The emotional state (i.e., anxiousness/negative emotional state) as well as fatigue will be measured with the help of ET technology based on the eye movement activity (below labeled as ET1) according to previously identified pattern analyses [116]. The distractions will be measured by ET based on the capturing specific fixations on the objects in the virtual environment (labeled as ET2) [68][69][70][71]. The Deception factor is represented by the conscious violation of traffic rules and could be identified based on a combination of ET1 and ET2 measures and activity log (e.g., driver saw speed sign but accelerated, considering also pupillometric measures to confirm consciousness rule violation) [108].…”