Voice Activated Migraine Assistant

Inspiration

Students with chronic migraines are often forced to miss large amounts of school, causing assignments to pile up. As most assignments are computer based, and migraines often have a light sensitivity component, it is incredibly hard to stay caught up while struggling through yet another migraine. Teachers often have little to no idea how to handle the rare case, sometimes even becoming convinced that the migraines are merely an excuse to avoid tests or schoolwork.

What it does

Our website seeks to help alleviate this problem, helping teachers understand and accomodate such individuals. For students experiencing chronic migraines, VAMA provides a dark, easy on the eyes, application for managing homework and due dates. Voice activated and screen reading features allow students to avoid the screen altogether, and the voices are optimized for gentleness on the ears. Teacher emails are stored alongside quick and flexible templates to allow students to communicate quickly and efficiently even on low brainpower. Integrations with printing and scanning make it easy for students to work out assignments on paper, minimizing interactions with screens. For teachers, VAMA provides a template for accommodating such students, educating them on what they need to provide to make it easy for their students to complete assignments asynchronously despite migraines.

How we built it

We built VAMA in Svelte, a lightweight and efficient alternative to heavier web frameworks such as React, which provides an extremely smooth user interface with incredibly low load tiems. For the backend, we used TypeScript, a version of JavaScript which provides typing functionality, vastly reducing development errors and aiding debug efforts. For database, we used Better-SQLite3, a battle-tested open-source file-based variant of SQL, which greatly simplifies the database interface for small-scale projects without sacrificing advanced features. We used the Web Speech API for the text-to-speech and speech-to-text features because it is widely available, cross-platform, and builtin to modern browsers.

Challenges we ran into

Our first and largest challenge was how to design a system which would actually be usable by individuals with chronic migraines. Not being affected ourselves, we had to rely on our knowledge of other's experiences and our own research for what would actually be helpful in such a situation. Our original concept merely involved extreme dark mode and printing functionality, but a friend's experience with migraines informed us that dark mode was not enough. Though it helps, a screen is still a screen. That is when speech recognition and screen reading entered the project, but still we had more to learn. Discussing the project with another friend, he informed us that even a voice can be grating, so we had to redesign our text to speech system to be gentle on the ears and filter out harsh sounds.

Accomplishments that we're proud of

Despite the difficulties we faced, we nonetheless succeeded in achieving a strong proof-of-concept for the voice recognition and screen reading portions of the project. Those parts were the largest question mark in the project, everything else is more mundane. We had experience in web development, but no experience with text-to-speech or speech-to-text, so we had mere hours to research and learn APIs on the fly. Our final project proved relatively robust, a strong achievement for our first efforts in the area.

What we learned

Building an application for individuals with a disability that we ourselves do not have opened our eyes to the difficulties faced by that portion of the population. The more we learned, the more we found out how little we knew. Our respect for those who struggle through with complications such as these increased steadily as we learned the lengths which they must go through to complete basic schoolwork. Our commitment to helping individuals with chronic migraines only grew as the project progressed, and we wish to help more people realize the difficulties which surround that condition, and other obscure conditions.

What's next for Voice Activated Migraine Assistant

The next steps for VAMA include expansion of the proof-of-concept features into full, usable components. The final goal for the printing and scnning features is to fully avoid the system dialogs if possible, allowing a fully voice controlled printing and scanning experience. Expanding the templates for the emails to be voice-fillable will make the process of notifying teachers far simpler and faster. Support for more languages will also allow VAMA to aid more people.