Sheild (She + Shield) – Total Safety, Indoors & Out

Inspiration

The idea for Sheild was inspired by the growing concern around women's physical safety—both in public spaces and private environments like hotel rooms or rented accommodations. While many apps offer basic SOS alerts, few provide real-time insights, while most don't cover safety risks in indoor situations like hidden surveillance threats. We wanted to build something intelligent, proactive, and discreet, giving women confidence whether they're walking down a street or checking into a room.

What it does

Sheild is a dual-mode, AI-powered safety app that provides real-time protection indoors and outdoors using AI environment detection. Visual and audio information about the surroundings environment will be processed and analysed to detect possible threats.

Toggle Modes

Auto or Manual Switching between indoor and outdoor modes using GPS + gyroscope data

Outdoor Mode (SafeWalk)

Uses AR-enhanced navigation with live camera overlay to show:
- Green Path: Safe zone (no recent threats)
- Red Path: Mark areas with high-risk patterns
- High Risk Zones: Detected threats nearby (based on data about the area such as crime reports, news feeds, and user reports, which is scraped in real-time)
If the user enters a High Risk Zone:
- Front and back cameras recording are activated automatically
- Voice recording starts until the user leaves the danger area
Live notification: based on environmental AI analysis, some safety factors such as light level, crowdedness, and notified to user in two ways:
- On-screen text: Alerts such as "Low Light Level! Be aware of your surroundings!", "Entering crowded area! Watch your personal belongings!", or "Unknown person ahead, be careful!"
- Audio notification using text to speech: These alerts will also be played, either through headphones or on speaker

Indoor Mode (SafeHaven)

Activates camera scanner to detect hidden surveillance devices
Uses IR reflection analysis and flicker detection to spot secret cameras
Designed for use in hotel rooms, public restrooms, and private rentals

How we built it

Design: Figma for mockups and intuitive UX

Challenges

Data accuracy: Crime data sources vary in format and reliability. Reliable data can be incorporated from Victorian Government crime reports.
Performance issues: Running AR + recording + scraping affected performance on low-end devices
Testing indoor features: Simulating realistic conditions for hidden cam detection was complex
Balancing UX & stealth: We had to design an app that’s functional but also discreet for high-risk scenarios

What we learned

Best practices for data privacy when dealing with sensitive information
The importance of empathetic UX design for vulnerable user groups
How to rapidly prototype and pivot under time pressure

What's next for Sheild (She + Shield)

Improve data reliability by partnering with law enforcement APIs and local NGOs
Add voice-activated SOS with custom safe-word triggers
Enhance indoor detection with thermal imaging support (via external devices or IR filters)