Elevator pitch Mental health and depression is a leading issue in the US. We created a device that will help track what may affect your mental health, along with extensive AI and non AI-powered tools to analyze it.

What is a PDA (no, not that kind!)

PDA stands for personal digital assistant, and if you haven’t heard of them, that’s because they were largely replaced by smartphones in the early 2010’s. But they are making a comeback. Phones today can cost hundreds of dollars, are addictive, and could potentially be damaging your mental health. And that’s what our product seeks to solve: mental health.

The Positivity PDA

The goal of the Positivity PDA is to help users pinpoint what could be negatively, or even positively affecting their mental health, while also promoting good habits. The purpose of the PDA is to collect data which will then be uploaded to our AI powered analysis website. On the website, users can easily correlate which activities are affecting their mood and get advice from the integrated Gemini LLM.

We also created a chrome extension that integrates with this website. It collects your youtube watching history and tracks how much time you have been wasting on unhealthy and non productive videos. The extension will use the Gemini API to get the video transcript and analyze if it is worth watching for you. It analyzes the productivity effects, the time wastage effects, and mental health effects. If it’s deemed unworthy, a popup will ask if you want to continue watching the video or not.

Collecting data

Throughout the day, the Positivity PDA prompts the user for two different scores: How well did you sleep? How was your day?

Additionally, we implemented three statistics the user can log: When they have unhealthy food, doomscroll, or drink caffeine.

The final method of data collection is the screen time tracking Chrome extension. This extension will track screen time throughout the day and communicate it back to the website to be analyzed. The extension has a bonus function of helping users be less distracted by YouTube videos (the PC version of doomscrolling), by rating the quality of the video prior to watching, using Gemini’s built in functionality.

Analyzing the data

Now that we have collected necessary data, we need to pinpoint potential stimuli affecting the user’s mood, and we have three methods of doing this: graphing the trends, statistical analysis, and finally the Gemini-powered analysis.

Graphing the data After the PDA is plugged in and uploads the data, it will be displayed on 3 graphs. Data per day, per week, and per month. With this, short term trends (such as having more sleep on weekends) and long term trends (falling into bad habits), can easily be seen. Statistical analysis: We used Chart.js to plot the data…..

How did we build it?

Hardware:

The PDA is controlled by an Arduino Nano inside a 3D printed case, along with 2 buttons and an E-Paper Display (EPD). A driver board with a level shifter interface between the Nano and the EPD. The whole thing is powered by a 9V battery.

Extension

We created a simple Chrome Extension that will monitor the youtube videos you watch. It will extract the video ID from the URL and use a library to get the transcript from the video. It then sends this to Gemini with a system prompt to analyze the content of the video and deem its time worthiness, and mental effects from this. The extension will then receive this information and show a popup to the user with this information with warnings if the score falls below a threshold. It also saves this data for the analysis.

Server + Frontend

The frontend get data from our API to retrieve the YT watching analysis. It then combines it with the data from the hardware to create a user behavioral profile of the user. This data is then sent to Gemini once again to create a final analysis for giving suggestions and feedback to the user explaining why they could be feeling the way they are and how they can change their lifestyle to improve.

Built With

Share this project:

Updates