Model Specification and Confidence Intervals for Voice Communication

There is an ongoing need for modeling voice communications in industrial applications, with system performance often depending on the accuracy of this information transfer. This article presents a case study using data from a human-in-the-loop experiment with a simulated flight environment conducted by the National Aeronautics and Space Administration (NASA) to investigate airborne spacing procedures. The interval management procedures during approach to an airport required a complex voice clearance issued by Air Traffic Control to a flight crew using radio communications. The time required for voice communication transfers is modeled, as is the time required for flight crews to complete data entry tasks. Commonly used reliability distributions are fit to the data, and the lognormal and log-logistic distributions are found to model the data reasonably well. Two analytical methods for calculating the confidence intervals for the lognormal mean are compared, and bootstrapping is used for log-logistic mean confidence intervals. Extensive investigation of outliers was performed to identify procedural anomalies. These initial results lead to design guidance for the phraseology used in air/ground communications.