ZeroBionic robotic sign language devices, 3D-printed from recycled plastic and priced at $350 each, are already in use in 78 schools across Kenya, built to bridge a gap that leaves most of the country’s deaf and hearing-impaired children outside formal education entirely.
Kenya has an estimated 300,000 children who are deaf or hearing impaired. According to RFI, just 20,000 of them were enrolled in schools as of 2024. The reasons are multiple, but one runs deep: Kenya has no standardised sign language, which makes it exceptionally difficult to teach subjects like mathematics or science to deaf students, and means most hearing teachers cannot communicate with them at all.
ZeroBionic was founded to address precisely that problem. The company, Business Daily Africa records, was co-founded in 2021 and is run from a small lab in Nairobi. Its core product is a robotic hand that listens to a teacher’s spoken words and translates them into signed gestures in real time, allowing hearing teachers to deliver lessons to deaf students without needing to learn sign language themselves.
Building ZeroBionic Robotic Sign Language From Scratch
The engineering challenge was only part of the problem. When Norah Kimathi and her team set about programming the translation layer, they discovered there was no unified Kenyan or African sign language dataset to work from. ‘No one else had built this on the continent,’ Kimathi told AFP. ‘We had to start from scratch.’
To build the dataset, ZeroBionic enlisted deaf teachers and experienced signers, who wore electrode-lined bodysuits (imported specifically for the purpose) to record the movements and patterns of signed speech. The resulting system now carries a 92% speech-to-sign accuracy rating, according to Kimathi. Getting there required far more effort than building the mechanical arm itself.
The hands are 3D-printed using recycled plastic litter, which keeps production costs down and gives the device a clear environmental rationale alongside its social one. Each unit is sold at $350 and is designed to last three years without maintenance. Those are numbers that matter in a market where cost and durability can determine whether a school actually uses a device or puts it in a cupboard.
What Comes After the First 78 Units
ZeroBionic has sold 78 units so far. That covers a small portion of what the need actually is, even measured against the 20,000 enrolled deaf students, let alone the much larger number still outside the school system. Kimathi has not presented the robotic hand as a complete solution, but as a first step in a longer programme.
The next phase is Africa One, a robotic upper torso that the company says would substantially expand both the accuracy and the vocabulary of African sign language translation. Moving from a single robotic hand to a full upper body would allow for more complex and nuanced signing, which matters most in technical subjects where precision is everything.
‘We know we’re going to put a smile on someone else’s face,’ Kimathi told AFP.
The difficulty of what ZeroBionic has done should not be understated. Building a functional sign language translation system in a context where no standardised dataset existed, no continental precedent had been set, and the electrode equipment had to be imported, required the team to solve problems that wealthier and better-resourced organisations had not solved first. The 92% accuracy figure and 78 sold units are early proof that the approach works. Whether the Africa One expansion secures the funding and scale to reach the remaining hundreds of thousands of children not yet in school is the open question that will define ZeroBionic’s next chapter.
