Skip to main content

Transatlantic Approaches for Automatic Speech Understanding in Air Traffic Management

Paper ID

ATM-2023-002

Conference

USA/Europe ATM R&D Seminar

Year

2023

Theme

Human Factors

Project Name

SESAR 2020 ER4 project HAAWAII, SESAR 2020 IR Wave 1 project PJ16 CWP HMI

Keywords:

air traffic control, automatic speech recognition, intent representation, natural language understanding, radio communications, semantic interpretation, semantic ontology

Authors

Hartmut Helmke, Shuo Chen, Oliver Ohneiser, Hunter Kopald, Robert M. Tarakan and Matthias Kleinert

DOI

Project Number

734141

Project Number

884287

Abstract

MITRE in the US and European stakeholders under the leadership of DLR have independently developed ontologies for representing the meaning of controller-pilot ATC (air traffic control) radio communications. With the intent of benefiting from each other’s work, and possibly harmonizing the two ontologies in the future, DLR and MITRE performed a structured comparison of the two ontologies. To explore how local differences in vocabulary and ATC procedures influence ontology design, both parties exchanged the transcripts of 100 ATC radio transmissions from their terminal airspaces. This paper summarizes similarities and differences of the ontologies from MITRE and Europe. Overall, despite a 12% difference in word level representation, 80% of the instructions from both data sets have common semantic representation in both ontologies.