Introduction
A team of Omani university students has developed an artificial intelligence-powered platform to make educational materials more accessible to learners with disabilities. By leveraging the latest advances in natural language processing and machine learning, the tool transforms traditional textbooks and digital content into audio files, braille-ready text, simplified summaries and sign-language videos. This initiative promises to break down barriers in education, offering students with visual, hearing or learning impairments a customized, more inclusive way to study.
Background
Education equality has long been hindered by a lack of tailored resources for students with disabilities. Conventional corrective measures—such as human-produced braille texts or recorded lectures—are labor-intensive, costly and slow to reach learners. Recognizing these challenges, a group of four computer science majors from Sultan Qaboos University partnered with educators and special-needs experts to harness AI for real-time content adaptation. Their goal: create a scalable, cost-effective solution to serve students across Oman and beyond.
Development of the AI Tool
1. Team Composition and Collaboration
– Four senior students in computer science and information technology.
– Academic supervision from the university’s Centre for Educational Technology.
– Regular consultation with teachers at local schools for the blind, deaf and students with dyslexia.
2. Technical Foundation
– Core engine built on an open-source transformer model fine-tuned with Arabic and English educational texts.
– Speech-to-text and text-to-speech modules powered by a state-of-the-art voice recognition API.
– A braille converter that outputs files compatible with standard embossers and refreshable braille displays.
– Integration of a sign-language avatar system that automatically generates short video clips for complex terminology.
3. Accessibility and Data Security
– User data is encrypted in transit and at rest, complying with international privacy standards.
– The platform requires minimal personal information: a username, password and profile indicating the user’s preferred accessibility mode.
– Regular security audits and ethical reviews ensure compliance with the university’s research standards.
Key Features
1. Multi-Format Content Conversion
– Upload PDF, Word or plain-text documents.
– Instant conversion to audio, braille and sign-language video.
– Option to download accessible versions or stream content directly from the browser.
2. Simplification and Highlighting
– Automatic summarization of long passages, highlighting key points and definitions.
– Adjustable reading levels: from basic vocabulary simplification to graduate-level paraphrasing.
3. Interactive Practice Modules
– Short quizzes generated from lesson content, with voice-prompt options.
– Real-time feedback on performance, with hints presented in the user’s chosen format.
Pilot Testing and Feedback
1. Initial Rollout
– Conducted at two special-needs schools in Muscat: the Al Noor Centre for the Blind and the Oman Institute for the Deaf.
– Over 60 students tested the platform during a four-week trial covering mathematics, science and language topics.
2. Observations and Outcomes
– Visually impaired students reported a 40 percent increase in study time, thanks to faster audio conversion.
– Deaf students found the sign-language videos helpful for grasping scientific terminology.
– Learners with dyslexia appreciated the simplified summaries and adjustable font sizes.
3. Testimonials
– “I used to wait weeks for braille-printed notes. Now I upload my files and get what I need in minutes,” said a 12-year-old student at Al Noor Centre.
– A sign-language interpreter noted, “The automatically generated clips are not perfect, but they cover most concepts accurately and save us hours of manual work.”
Future Prospects
1. National Expansion
– Plans to partner with the Ministry of Education to integrate the tool into public-school e-learning systems.
– Training workshops for teachers and school IT staff are scheduled later this year.
2. Feature Enhancements
– Improved accuracy of sign-language sequences through deep-learning retraining on larger video datasets.
– A mobile app version enabling offline access for students in remote areas.
3. Commercial and Global Reach
– Discussions underway with a local startup incubator for seed funding.
– Potential to license the platform to educational institutions in the Gulf Cooperation Council and North Africa.
Three Key Takeaways
• This is the first AI-driven educational accessibility tool developed entirely by Omani students, demonstrating local talent in cutting-edge technology.
• By converting standard educational content into multiple accessible formats on demand, the platform addresses a critical gap in resources for students with visual, hearing and learning impairments.
• Early trials show significant improvements in study efficiency and comprehension, paving the way for national and regional adoption.
Frequently Asked Questions
1. What formats does the tool support?
The platform accepts PDF, Word and plain-text files, converting them into audio (MP3), braille-ready text (BRF), simplified summaries (PDF) and short sign-language videos (MP4).
2. Who can use the AI tool?
Any student or educator with a registered account can use it. While initially focused on K-12 learners with disabilities, the team plans to extend support to university students and adult learners.
3. How can schools integrate this solution?
Schools can apply for a free pilot program through the university’s Centre for Educational Technology. Upon approval, administrators receive training materials and ongoing technical support, with the option to subscribe to a low-cost premium plan for advanced features.
