evidoria

← Back to browse

Good practice Imported

TransportSign — Real-Time AI Sign-Language Avatars for Serbian Transit Announcements

Serbia · Belgrade · See the Serbia profile · See the Belgrade profile

Top 86% 36/100 · Ask Evidence Copilot about this practice

Belgrade startup SignAvatar's TransportSign turns live station/airport announcements into sign-language avatar video on public screens, live in Belgrade since 2024. WSA- and OECD-documented, but evidence is still just the company's own small self-reported survey.

TransportSign — Real-Time AI Sign-Language Avatars for Serbian Transit Announcements

Details

Promoter
SignAvatar, in partnership with Railways of Serbia
Period
2024–present
Keywords
transport, accessibility, artificial intelligence, sign language

Description

TransportSign is the first product of SignAvatar, an AI startup founded by a small team of Belgrade engineers. It uses AI transformer models and a hyper-realistic digital avatar to convert live public-address audio — station and gate announcements, delays, safety notices — into sign-language video shown on screens in real time, alongside text translation. The system supports International Sign as well as national sign languages including Serbian Sign Language (SZJ) and Macedonian Sign Language.
Under a licensing agreement with Railways of Serbia, TransportSign has been running at Belgrade Centre Railway Station and at Belgrade Airport since 2024, addressing a gap that affects an estimated 460 million Deaf people worldwide who rely on audio-only public announcements. The project was named a 2024 WSA (World Summit Award) European Young Innovators Award winner and documented as an innovation case study by the OECD's Observatory of Public Sector Innovation (OPSI).
Evidence to date is early-stage: the company's own account cites that 'over 85% of surveyed deaf individuals reported heightened travel safety,' without a disclosed sample size, sampling method or independent verification. The deployment remains limited to one station and one airport, with stated — not yet contracted — interest in expanding to two further stations, and no independent, peer-reviewed study of transcription accuracy or comprehension outcomes has been published.

Read the full analysis: https://oecd-opsi.org/innovations/signavatar-ai-speech-to-signlangauge-translator/

Implementation

Implementation detail (cost, timeline, staffing, conditions for success) is not yet available for this practice.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful