Inspiration Today’s students are struggling with mobile addiction due to constant notifications, social media, and games. This affects their studies, focus, mental health, and sleep. We wanted to build a smart companion that understands students’ behavior and helps them use technology responsibly without taking away their freedom.
What it does Our app FocusGuard – AI-Based Mobile Addiction Controller:
- Tracks mobile usage behavior automatically
- Predicts addiction level using machine learning
- Blocks and limits distracting apps during study hours
- Sends night-usage alerts to prevent sleep disturbance
- Generates weekly performance reports
- Provides rewards for healthy digital habits
- Allows guardians/counsellors to monitor improvement
It promotes digital discipline in a friendly and intelligent way.
How we built it We developed:
*Mobile App Layer — Kotlin / Android Studio
- Usage Tracking — Android UsageStats API
- ML Model — Python, Scikit-Learn, TensorFlow Lite
- Backend & Auth — Firebase Authentication + Firestore
- Dashboards — Real-time graphs & weekly analytics
- UI/UX — Modern design using Material Components
We first created a rule-based model, then upgraded to a TFLite AI prediction system for accuracy.
Challenges we ran into Handling
*Background usage tracking without draining battery *Managing user privacy & permissions *Designing accurate addiction scoring logic *Making communication between app & ML smooth
- UI testing across screen sizes
- Time constraints — integrating all features before demo
Every challenge helped us get better at problem-solving.
Accomplishments that we're proud of
Our app runs 100% automatically — no manual control needed *We successfully deployed AI on mobile using TensorFlow Lite *Real-time dashboard shows measurable improvement
- Guardian monitoring without privacy invasion
- Clean UI experience with smart alerts & focus mode
We built something that can genuinely improve students’ lives
What we learned How to integrate AI into mobile apps
- How to collect & process real-world behavior data responsibly *Balancing usability with restrictions
- Full development cycle — design → build → test → deploy
- Team collaboration using version control & task sharing
This project made us future-ready developers
What's next for AI-BASED MOBILE ADDICTION CONTROLLER
- Voice-based reminders & motivation assistant
- Personalized focus coaching using deep learning
- Geofencing (auto-focus when entering school/college)
- Integration with smartwatch (heart rate stress analysis)
- Community challenges — friends can compete on discipline
- Fraud detection if user tries uninstalling or disabling app
We aim to launch this as a real app for schools & colleges to support student well-being.
Log in or sign up for Devpost to join the conversation.