A graduate of the University of KwaZulu-Natal (UKZN) has created an artificial intelligence-based system capable of translating South African Sign Language (SASL) into speech in real time. The goal of this development is to help people with hearing impairments overcome communication barriers.
Akhil Hansraj developed the SASL-to-Speech Translation System as part of his undergraduate engineering degree project in Computer Engineering. This system relies on computer vision and utilizes a custom-developed AI model alongside an established machine learning framework to track body landmarks and recognize signs in real time.
The system was trained on a specialized dataset pertaining to SASL and can identify individual letters and words before converting the interpreted signs into speech using a text-to-speech model.
Overcoming Challenges
According to Hansraj's academic supervisor, Dr. Sulaiman Patel, this project was one of the first assistive technology projects in the Electrical, Electronic, and Computer Engineering (EECE) discipline at UKZN specifically focused on the South African sign language dialect. It is worth noting that SASL was recently recognized as the eleventh official language of South Africa, while many international sign language translation systems focus on American Sign Language.
For Hansraj, this project held deep personal motivation, as both of his parents were born with hearing impairments and faced communication difficulties in daily situations. Hansraj noted: 'My parents face communication barriers in everyday settings such as banks, government offices, and municipal offices.' He added that he and his sister often had to accompany their parents to speak on their behalf.
Hansraj believes that a fully functional version of the technology will allow people to communicate more independently. He stated: 'A production-ready system will give these people the ability to communicate more easily and effectively, without the need for the other party to learn SASL.'
Despite this outstanding achievement, Hansraj's path to graduation was not easy. Patel reported that in the second semester of 2025, Hansraj experienced mental health issues and required an extension for his first project submission. With Patel's support, including referrals to student counseling and regular check-ins, Hansraj was able to regain momentum on the project. His efforts ultimately earned him high praise from EECE for his final project.
Hansraj also shared that he enjoyed the design modules the most during his studies because they allowed him to apply theoretical knowledge to practical, meaningful projects. He also named the 'Network Communications' and 'Internet Engineering' modules, taught by Professor Tahmid Qazi, as his favorites. With one final module in 2026 and facing financial difficulties, Hansraj was recommended by Qazi to a local company for an interview. He secured the job and balanced it with full-time studies.
Future Plans for Technology Development
After obtaining his bachelor's degree in engineering, Hansraj intends to continue developing his sign language translation project. The main focus will be on improving the system's ability to recognize words and generate more coherent and fluid speech. In the long term, he plans to build a career in software, network technologies, or artificial intelligence and machine learning.
Hansraj expressed gratitude to his parents, uncle, sister, Patel, friend Amil Maharaj, and his roommate, girlfriend, and university colleagues for their support throughout the journey. Besides his studies, he enjoys gaming, keeping up with technological developments, and working on his own programming projects. He also thanks UKZN for the knowledge gained and the opportunity to experience university life.