Digital Transformation
The ATCO2 project aims to develop a single platform that can collect, store, process and share real-world air traffic control voice communications using deep learning methods. These machine learning solutions will facilitate the development of air traffic control technologies.
The project aims to access data from certified ADS-B datalinks integrated with surveillance technology, which is provided directly by air traffic controllers and made available by air navigation service providers.
The project is built around a robust platform based on an existing and widely used solution from the partner OpenSky Network, ensuring its long-term sustainability. To enable data collection, voice communications and time/position data are being stored and pre-processed alongside other aircraft-related information. The project targets voice commands issued by air traffic controllers and readback confirmations provided by pilots. In addition to the transmitted data, ATCO2 will have access to voice recordings from air navigation service providers (e.g. Austrocontrol). Automatic segmentation has also been implemented and integrated with robust automatic speech recognition systems to transcribe voice communications automatically, thanks to machine learning algorithms capable of iterative improvement and manual post-editing.
In line with the CleanSky2 programme, the project will contribute significantly to community building by strengthening the existing ‘OpenSky Network’. Users will be incentivised to upload and potentially pre-transcribe data in order to access other resources and automatic transcriptions. The project also considers legal and ethical issues relating to privacy, personal data, data security and related matters.
Romagna Tech S.C.p.A.
VAT number and tax code 02327720401 - REA no. 258255
Fully paid-up share capital € 611.500,00
Registered office C.so Garibaldi, 49 47121 Forlì (FC)
T +39.0543.32087
