Targeting Decisions Using Multiple Imaging Sensors: Operator Performance and Calibration

Abstract

Two experiments investigate target identification performance and operator calibration (i.e., the ability to evaluate the accuracy of one's own performance) using sensor imagery from either a single information source or two separate sources. The experimental task required subjects to identify a target ship among two distractors using either simulated forward-looking infrared (FLIR) imagery, simulated range only radar (ROR) imagery or both information sources presented simultaneously. Relative to the single-sensor condition, performance in the dual-sensor condition was either enhanced or degraded depending on the quality of the information presented on each sensor. Operator calibration was found to be poor, with operators consistently underestimating the accuracy of their target identification performance. Similar patterns of performance and calibration were found for both non-aviators (Experiment 1) and experienced aviators (Experiment 2). The finding that operators were often unable to integrate effectively two sources of sensor information implies that it might not be advisable to provide pilots with all of the information available in a multi-sensor suite unless additional cues are provided to aid in the integration of the information. Poor operator calibration could negatively impact targeting by leading to inappropriate delays (underconfidence) or misidentifications (overconfidence). Suggestions for improvement of operator information integration and calibration are offered.