The innovative Speech2Speech translator of DB Systel and its cooperation partner, DB Netz AG, for direct communication in cross-border rail traffic is based on combinable modules that convert voice into text, translate it into the foreign language and finally reproduce the translated message with a natural voice. The AI-based application enables almost natural communication between train drivers and dispatchers in their native languages without having to master the foreign language of the other - all in compliance with the established safety standards of international railway operations.
Your Benefits with KITT:
- KITT offers reliable translation beyond language barriers - especially "in railway language" - not only for Deutsche Bahn,
- but also for Infrastructure Managers and all Rus across Europe
- KITT reduces the cost of language training by at least 50%
- KITT helps to reduce the demands on operating personnel and enables their more flexible assignment
- KITT enables the recruitment of staff with minor language skills
How does our translation tool for cross-border rail transport work? Artificial intelligence explained simply from the DB Systel series of explanatory videos!
A Team of Speech Experts:
Spoken language is the human ability of putting thoughts into words. This particularly important ability in human communication consists of many complex processes that also include body language and facial expressions. In railway operations and other areas of life, these elements are absent and one has to rely exclusively on spoken language.
At DB Systel GmbH, we use and combine the latest language technologies on the market to enable multilingual communication in railway operations with established safety guidelines. Our team of computer linguists will advise you and your company to design and develop customised solutions in line with your needs.
We develop, together with the language management of the DB Group, unique, brand-compliant and intelligent applications for you and operate them with the language technology that suits you best.
SINUS - Voice-Assisted Maintenance and Inspection
The MVP for the SINUS voice assistant was successfully implemented in cooperation with our partners TechEx and DB Fahrzeuginstandhaltung in Dessau. The SALT team trained the voice models for the application and tested them with the maintenance staff on site in Dessau. SINUS uses a range of innovative technologies such as speech recognition, dialogue management, natural language understanding, voice playback and Knowledge Graph to ensure flawless input by means of voice control for train evaluation. This makes evaluation easier for maintenance staff. At the same time, it is made more cost-efficient.
The results are impressive. Feedback from the staff is very positive as they are able to carry out hands-free evaluation. The Group partners were also very satisfied with the ergonomics and practicality of the headsets. SINUS facilitates thorough evaluation, which in turn significantly improves the completeness of results. Using the assistant makes evaluation more efficient.
In future, we are planning to continue scaling the assistant to other organisations within the company.