Speech Inputs To Surface Safety Logic Systems

In Air Traffic Management, radio communications remain the primary method of information transfer between controllers and pilots. As such, controller-pilot voice communications comprise a wealth of context and intent information that could potentially be leveraged to augment existing surveillance and flight data sources to the benefit of current surface safety logic systems. Automatic speech recognition provides a means of mining the voice communications for information of interest. An additional benefit of integrating automatic speech recognition with a safety logic system is the potential speech recognition performance improvement that the incorporation of context information could enable. This paper describes how surface safety logic and automatic speech recognition can be used together in a mutually beneficial relationship. A case study safety logic system is presented to describe how speech inputs can enhance safety logic performance, and preliminary speech recognition performance analysis is summarized to demonstrate the benefit enabled by using context information to augment automatic speech recognition on Tower controller voice transmissions.