In Kenya, a young team of innovators is applying robotics to overcome one of the obstacles faced by hearing-impaired students: access to education in sign language.
Although this technology is still in the development stage, its creators believe that in the future it could make technical subjects more accessible for students with hearing loss.
In the classroom in Nairobi, communication largely occurs through gestures. However, for deaf students, mastering complex disciplines presents a particular challenge, especially when the vocabulary needed to explain mathematics, science, or technology is absent from the signs they know.
Eva Avino, a teacher at the secondary school for the deaf Kasarani Treeside Secondary School for the Deaf, noted: 'I teach physics, and most of the time we understand that the concepts we teach and the way I can express the concept become problematic at some point due to the signs and vocabulary required to express that concept.'
This is the problem that the Kenyan startup ZeroBionic is trying to solve. In their laboratory in Nairobi, engineers are training machines to communicate in sign language.
For co-founder Nora Kimati, the idea started with something much simpler. Kimati said: 'We started with basic robotic arms. Since sign language is entirely based on hands, we decided to start with hands and see how it would go. Then we received feedback that a lot in sign language depends on it.'
She continued: 'For example, performing this means more than five different things in a specific sign language, depending on your facial expressions. That's when we decided to build a full humanoid, including the face, including the eyes, including everything, to capture all the non-verbal cues that are currently present.'
However, teaching a machine signs requires more than just hardware; it requires data. When ZeroBionic began searching for African sign language datasets, the team reported that there was very little material available for work.
In the lab, every movement is captured, recorded, and converted into data that can train the system. ZeroBionic claims that the current task is to make the technology smarter, more accurate, and more useful in the real world. For students watching this technology develop, it symbolizes something more than just a machine.