Speech technology for railway traffic – far more than digital announcement voices!

To improve their passenger information systems’ quality and economic efficiency, the Swiss railway services SBB (Schweizerische Bundesbahnen) published a tender. Aristech won the competition by impressing the SSB with our text-to-speech (TTS) systems’ quality. The SSB now plans to equip all train stations and stops in the entire rail network with Aristech’s TTS technology in four languages (German, English, French, Italian) within the next years.

These modernization measures are based on our professional systems, which generate the desired announcements in consistently comprehensible and pleasant voices. All texts entered by the train station in charge are immediately converted to audio – in four languages, grammatically correct, with a maximum of flexibility in the event of changes.

The voice’s recognizability has the same effect as a font logo: It is a part of a company’s corporate identity. Our corporate voices are based on human voices. A voice talent records many different sentences as evenly and naturally as possible. This ensures that longer text parts still fit together smoothly. We then edit the recordings and use the cleaned, straightened and optimized material to create a digital voice in elaborate automated and manual processes. Finally, the maximally optimized audio material is used to create the digital voice, which can dub any input text and can thus be used for a wide variety of applications.

For this we benefit from broad language technology know-how, machine learning and the use of semantic AI.