The Education Access Project was developed through multiple iterations, with each stage refining both the hardware and the content creation process. Rather than beginning with a fully developed system, the project started as a simple proof of concept intended to answer a single question:
"Could low-cost, offline audio lessons become a practical supplement to STEM education in rural schools?"
The answer to that question shaped every design decision that came after.
Before selecting hardware or developing software, the project needed clear objectives.
The system needed to:
Operate without an internet connection after deployment.
Be affordable enough for schools to adopt at scale.
Support bilingual educational content.
Store an entire semester's worth of lessons.
Provide sufficient battery life for regular student use.
Be simple enough for students to navigate independently.
Allow new lessons to be produced efficiently.
These requirements shaped every engineering decision made during development.
The first prototype was built using a Raspberry Pi.
The Raspberry Pi was chosen because it provided an inexpensive platform for testing the core concept without requiring custom hardware. Audio files were stored locally, and a simple text-to-speech workflow was used to generate sample lessons. This prototype demonstrated that offline audio could effectively deliver educational content and confirmed that the overall concept was technically feasible.
Although the prototype achieved its primary goal, it also highlighted several limitations.
The Raspberry Pi offered considerably more computing power than the project required. Since the final system only needed to store and play audio files, much of the hardware remained unused while significantly increasing the overall cost of each device.
This prototype demonstrated that the concept was viable, but it also showed that a different hardware platform would be necessary for large-scale deployment.
After discussing the prototype with Mr. Srivastava, one requirement became clear: the system needed to remain affordable enough for schools to realistically purchase multiple devices.
The target hardware cost was approximately ₹1,200 (about $12.50 USD) or less per unit.
With this constraint in mind, several alternatives were evaluated before finally landing on dedicated MP3 players, sourced through Amazon. This decision was due to the multitude of advantages they had over the Raspberry Pi. Lower costs, long battery lives, and reliable playback were major reasons, as well as the ease of use and upkeep that would be required over the Raspberry Pi.
Most importantly, the MP3 players focused exclusively on the functionality students actually needed. Rather than functioning as miniature computers, they served as straightforward educational devices that students could operate with minimal instruction.
The final hardware cost was approximately $11 USD per device, meeting the project's affordability target while maintaining adequate storage capacity and battery performance.
Choosing how lessons would be distributed was another important design decision.
The project standardized on the MP3 format for every lesson because it provided an effective balance between audio quality, storage efficiency, and compatibility.
Using MP3 files allowed hundreds of lessons to fit comfortably on each device while maintaining clear speech quality. Since MP3 is a widely supported format, lessons can also be transferred easily between devices or archived for future use.
Keeping every lesson in the same format simplified both content production and deployment.