LOCKIN!

Inspiration

LOCKIN! 🔒 was inspired by us seeing first year Engineering Sciences students at UofT scrolling short reels on social media during classes and study sessions. Observing how easy it was for students to lose focus and get distracted by these reels highlighted the necessity for a tool that not only promotes focus but makes studying engaging and productive. We realized that traditional study methods were too boring and lacked interactive elements that could keep students engaged for longer periods. Thus, we decided to create an innovative platform that blends gamification with productivity tools to enhance the studying experience.

What it does

LOCKIN! 🔒 is an AI-powered study platform that transforms studying into a gamified and engaging experience while keeping users focused. Key features include:

Gamified Study Experience: Points-based system and leaderboard to inspire healthy competition. Users can form groups, work together, and earn points for staying on task and achieving goals.

AI Attention Tracking: Advanced machine learning models monitor user attention levels. If focus drops, LOCKIN! steps in to prompt users back on track.

Custom Practice Material Generation: AI integrated with PerplexityAI generates practice questions from PDFs like lecture notes or syllabi, tailoring the learning experience.

Interactive Feedback Mechanisms: When focus dips, the hardware elements provide real-time feedback. An Arduino-powered water squirt gun gives amusing reminders, and a camera-equipped pig robot monitors workspace activity while playing motivational sounds, including the iconic Squid Game doll alert.

How we built it 🛠️

For the frontend, to reduce networking issues and compatibility issues, we used Streamlit to write a seamless backend with the frontend. Individual classes were used to simulate models so that people can work in groups with a leaderboard of which group is the most focused on their task with points given to teams with SupaBase. Then, we used Tensorflow model + OpenCV to track the attention of the user. LOCKIN! also determines whether you are focusing on reading a document in an adequate amount of time and we used PerplexityAI to create practice questions based on a PDF document that the user provides. This PDF document could be a course syllabus or lecture notes.

When a user is not focused, a squirt gun controlled by an Arduino with servo motors are used to spray some sense into them. We also have a cute pig that has a camera to monitor your activity to see if you are actually doing work or not! It also makes the squid game doll sound to really entice them to get back to work.

To secure this app, of course we had to use Auth0 to ensure MFA and login using Google feature. This was to ensure user data is not, and there are no people trying to impersonate people so that they aren't locked in.

Challenges we ran into

Frontend-Backend Integration: Even though Streamlit streamlined the development process, integrating complex backend logic with real-time AI tasks presented compatibility and versioning issues with compatibility issues (especially between mac and windows.

Hardware Implementation: Setting up the squirt gun and camera system with Arduino and synchronizing it with AI tracking mechanisms proved to be challenging. Precision in controlling these devices required additional calibration and troubleshooting.

Real-Time AI Processing: Ensuring accurate and low-latency attention tracking with TensorFlow and OpenCV required careful optimization, especially for seamless performance on varying hardware setups.

Accomplishments that we're proud of

Successfully integrating AI for real-time attention tracking to enhance user accountability. Creating a cohesive platform that combines productivity, gamification, and interactive elements. Overcoming technical challenges to design hardware-based features, like the water squirt gun and camera-equipped robot, for boosting engagement and maintaining focus. We also built a system that generates personalized practice content from course materials, bridging the gap between passive study and active practice.

What we learned

We gained hands-on experience with machine learning model optimization and real-time processing using TensorFlow and OpenCV. Designing gamified study experiences taught us how to balance functionality with fun, ensuring users are both productive and entertained. Combining Arduino with AI systems expanded our understanding of how to connect software and hardware in creative ways. Furthermore, 3 of us were beginners, so learning how to code and use Streamlit and building a full fledge app was a new experience.

What's next for LOCKIN!

We want to integrate with productivity tools by connecting with platforms like Google Workspace, Notion, and Microsoft OneNote for seamless workflow integration. Emotion detection to gauge stress levels and adapt feedback accordingly. We also want to further enhance customization in practice material creation using advanced natural language processing models.