Imagine you're in a meeting and one participant communicates in sign language while another speaks — and nobody can follow each other. SignON built a smartphone app that watches someone signing on camera, translates it into spoken or written language, and works the other way around too. Think of it like Google Translate, but for sign language — covering 5 sign languages and 4 spoken languages across Europe. The heavy processing runs in the cloud while your phone stays a lightweight window into the conversation.
Deaf and hard-of-hearing people face daily communication barriers with hearing colleagues, customers, and service providers. Hiring human sign language interpreters is expensive, hard to schedule, and impossible to scale for everyday interactions. Businesses in healthcare, education, and customer service lose engagement and face accessibility compliance risks when they cannot bridge this gap.
A smartphone app that translates between sign language video and spoken/written language, backed by a cloud-based processing system. Concrete outputs include: a final mobile application (with code), an open translation system released in 3 iterations, NLP processing pipelines, a 3D virtual character (avatar) for sign language output, and an interactive co-creation platform for learning from user input — 50 deliverables in total.
If you are a hospital or clinic dealing with deaf patients who struggle to communicate symptoms to doctors and nurses — this project developed a mobile translation app covering 5 sign languages that converts signing on camera into spoken or written language in real time. It can reduce miscommunication risk in clinical settings and cut the cost of hiring on-site interpreters for routine appointments.
If you are a customer service provider dealing with accessibility compliance and excluding deaf customers from support channels — this project developed an open communication system with sign language recognition and speech processing that runs on standard smartphones. With 16 consortium partners refining it across 5 countries, the technology is designed for real-world multilingual environments.
If you are an education provider dealing with the challenge of making lectures and online courses accessible to deaf and hard-of-hearing students — this project built a cloud-based translation service plus a 3D virtual character (avatar) that can present content in sign language. The co-creation platform allows continuous improvement from user feedback, covering Irish, British, Dutch, Flemish and Spanish sign languages.
SignON was developed as a free, open application and open-source communication system. The core code and mobile app are available as demonstrators. Integration costs would depend on customization, language additions, and cloud infrastructure for the computationally intensive backend.
The architecture separates the lightweight mobile interface from cloud-based processing where computationally intensive tasks run. This design supports scaling, though the project tested with community user groups across 5 countries rather than at enterprise volume. Additional cloud infrastructure planning would be needed for large-scale deployment.
The project was funded as a Research and Innovation Action (RIA) led by Dublin City University with 16 partners. The final release of the Open SignON system is described as an open communication tool. Specific licensing terms should be confirmed with the coordinator, but the stated goal was wide uptake and open availability.
It currently covers Irish, British, Dutch, Flemish and Spanish sign languages, plus English, Irish, Dutch and Spanish spoken languages. The system incorporates machine learning capabilities specifically designed for learning new sign, written and spoken languages through user feedback and adaptation.
The project went through multiple development cycles — first, intermediate, and final releases of both the mobile app and the translation system — with quality assessment by deaf and hard-of-hearing communities. Based on available project data, specific accuracy percentages are not published in the objective, but the user-centered co-design approach aimed to validate real-world acceptance.
The mobile application runs on standard modern smartphones — no special hardware needed. The phone captures video of signing and handles the interface, while the cloud backend processes the computationally intensive translation tasks. This keeps device requirements minimal.
The project aligns directly with EU goals for inclusive digital content and fair access to information. While the project itself is a research output rather than a certified product, it was designed with European accessibility needs in mind and could support compliance with the European Accessibility Act requirements.
The consortium of 16 partners across 5 countries is heavily academic — 9 universities and only 2 industry partners (12% industry ratio), with 2 SMEs. Dublin City University in Ireland leads the project. This composition is typical for a research-stage technology: strong on scientific depth and multilingual expertise, but lighter on commercial deployment experience. A business looking to adopt this would likely need to work directly with the university coordinator or the 2 industry partners who understand productization. The 5-country spread (Belgium, Switzerland, Spain, Ireland, Netherlands) gives good coverage of Western European sign language variants.
Dublin City University (Ireland) — reach out through their research office or the project website contact page
Want to integrate sign language translation into your product or service? We can connect you with the SignON research team and help evaluate fit for your use case.
