Development of real-time performance metrics for manually-guided spacecraft operations

Real-time performance metrics were developed to quantify workload, situational awareness, and manual task performance for use as visual feedback to pilots of aerospace vehicles. Results from prior lunar lander experiments were replicated and extended to provide insights for the development of real-time metrics. Automated Speech Recognition (ASR) was employed to detect verbal callouts and measure subjects' situational awareness. ASR achieved an F-measure of 0.8, despite large variances within subjects. A one-dimensional manual tracking task and simple instructor-model visual feedback scheme was developed. Experiments showed that for this low-complexity task, visual feedback did not change subject performance, but did increase the subjects' measured workload.