Study in application of natural language processing in maritime communications

Abstract

Recent research on maritime accidents done at the School of Nautical Studies at the University of A Coruña (Spain) shows that nearly 20 per cent of accidents in maritime settings have been due to, among other causes, communication problems derived from the lack or misuse of a common language. Moreover, automatic speech translation, a technology that combines speech recognition and automatic translation, has been for more than a decade the focus of research as a tool for improving communication in different settings. Similarly, this technology can be applied, to a greater or lesser extent, in communicative processes that take place in the maritime workplace so as to minimise problems stemming from multilingual environments, especially those in which communication takes place via the use of radio devices, in which the inclusion of an automatic translator could enable two people of different nationalities, for example, to communicate with each other in their own native languages. This article describes a system of this type whose feasibility is being studied at the Universidad de A Coruña, under the auspices of a university-financed project called "Language Industries Applied in the Maritime Sector". What this project seeks, therefore, is to study the legal and technical possibilities, as well as the commercial suitability of developing an automatic translator for oral communications in the maritime sector, and in such a case, to establish the foundations for such a development, as well as to study the implications that the use of a device with such features would have on maritime safety. Likewise, this project seeks to develop the simulator for learning and using maritime phrases and for the creation of the corresponding curricular material.