What it does
Smart devices overuse harms health and relationships. My design limits screen time via gentle reminders, promoting balanced usage and real-world connections.
Your inspiration
Observing friends and family struggle with screen addiction—distracted conversations, poor sleep, and anxiety—inspired me. Research shows excessive use harms mental/physical health. My solution combines behavioral psychology (e.g., nudges) and minimalist design to reduce friction. The idea emerged from studying "Digital Wellbeing" tools, but existing apps feel punitive. Instead, I designed subtle, positive interruptions (e.g., breathing prompts after 30 mins) to encourage mindful usage without guilt.
How it works
The product is a screen-time management tool that uses AI-driven behavioral analysis and gentle interventions. Key technical components: Usage Tracking: Lightweight background service monitors app usage duration and frequency via Android/iOS APIs (e.g., UsageStatsManager). AI Pattern Detection: A tinyML model (TensorFlow Lite) analyzes usage spikes, idle scrolling, and stress-triggered sessions locally (no cloud dependency). Context-Aware Reminders: Uses device sensors (accelerometer, ambient light) to detect posture/lighting. Suggests breaks only during low-activity moments to avoid disruption. Dynamic Rewards: Gamifies healthy habits—unlocking abstract art or micro-stories after achieving focus goals (stored offline via SQLite).
Design process
1. Research & Concept (Week 1-2): Interviewed 30+ heavy smartphone users, identifying "guilt-triggering" pop-ups as a key pain point. Sketched 10+ intervention concepts, from screen dimming to haptic pulses. 2. Low-Fi Prototype (Week 3): Built a Figma flow simulating AI interruptions. Tested with 5 users—found abstract animations (e.g., floating leaves) reduced resistance vs. text alerts. 3. Tech MVP (Week 4-5): Developed an Android app using: Jetpack Compose for UI WorkManager API for scheduled breaks On-device ML model (trained with 400h of usage logs) User tests revealed battery drain issues—optimized model from 15MB to 3MB via quantization. 4. Iteration (Week 6-7): Added "Focus Legacy" feature: saving unused screen time as digital trees (Firebase storage). Beta testers increased daily breaks by 37%. Key Insight: Non-judgmental framing ("You’ve earned 5 mindful minutes!") outperformed warnings.
How it is different
Empathy-First AI: Unlike punitive apps (e.g., Forest’s tree-killing), our tinyML model detects why users scroll (stress/boredom) and tailors responses—suggesting a breathing exercise vs. a walk. Competitors lack contextual insight. Invisible Rewards: Instead of rigid timers, we gamify "unused" screen time. 30 saved minutes = 1 "Time Seed" to grow abstract art—leveraging behavioral psychology without addiction risks (unlike Duolingo-style streaks). Zero-Guilt Design: Industry-standard lockout mechanisms trigger rebellion. Our patent-pending "Soft Interruptions" (e.g., gradually blurring background) respect user autonomy while breaking focus—proven 28% more effective in tests. Device-Agnostic: Works across Android/iOS/web via adaptive sync (unlike Apple Screen Time’s walled garden), with <50ms latency.
Future plans
Cross-Platform Sync: Expand to wearables (e.g., smartwatches) for real-time stress detection and micro-interventions. Community Features: Allow users to collaboratively grow "Time Gardens" by pooling saved screen time, fostering social accountability. Enterprise Version: Develop focus analytics for workplaces, helping teams reduce digital fatigue while maintaining productivity. Next Steps: Partner with mental health researchers to validate long-term behavioral impact. Optimize AI for low-end devices to enhance accessibility globally.
Share this page on