AegisPath

A safe route navigation system.

Comment

Inspiration

Every day, millions of people use navigation apps that optimize for the fastest route, but not necessarily the safest one. For women, students, late-night commuters, and anyone traveling alone, a route that saves a few minutes may come at the cost of poor lighting, isolated stretches, or limited access to emergency services.

We wanted to build a navigation experience that prioritizes personal safety as a first-class factor rather than an afterthought. AegisPath was created to answer a simple question:

"What if maps could guide you not just to your destination, but also help you get there more safely?"

What it does

AegisPath is an AI-enhanced safety-focused navigation platform that recommends safer routes instead of simply recommending the fastest ones.

The platform:

  • Analyzes routes using real-world map intelligence from OpenStreetMap
  • Evaluates crowd density indicators, lighting confidence, isolation risk, and emergency accessibility
  • Compares multiple route options and ranks them by safety
  • Provides route-specific safety breakdowns and risk analysis
  • Offers emergency SOS functionality with location sharing
  • Supports shake-to-SOS activation for rapid emergency response
  • Provides nearby emergency service awareness during navigation

To make safety information more understandable, Gemini acts as an AI Safety Copilot that interprets route intelligence and generates personalized safety insights, contextual warnings, and actionable recommendations for travelers.

How we built it

We built AegisPath using:

  • React Native
  • Expo
  • React Navigation
  • Zustand
  • React Native Maps
  • OpenStreetMap
  • Overpass API
  • OSRM Routing Engine
  • Gemini API

Our routing pipeline first retrieves route alternatives using OSRM. We then enrich those routes with environmental intelligence gathered from OpenStreetMap and Overpass API data.

A custom safety scoring engine evaluates factors such as:

  • Lighting confidence
  • Crowd activity indicators
  • Isolation risk
  • Emergency accessibility
  • Nearby points of interest

The safest route is then selected and presented to the user alongside detailed safety metrics.

Gemini is integrated as an AI reasoning layer. Rather than calculating route scores, Gemini interprets route intelligence and transforms technical safety data into understandable, personalized guidance and situational awareness recommendations.

Challenges we ran into

One of the biggest challenges was obtaining meaningful safety signals from publicly available map data.

Unlike traffic or travel time, safety is not a single measurable metric. We had to combine multiple indicators such as road types, nearby establishments, emergency services, and environmental context to estimate relative safety.

Another challenge was ensuring that AI complemented the safety engine instead of replacing it. We wanted Gemini to provide human-friendly reasoning and contextual guidance while keeping route scoring deterministic, transparent, and reliable.

Balancing route safety with practical travel times was also a key design challenge, as users often need safer routes without significantly increasing travel duration.

Accomplishments that we're proud of

  • Built a complete safety-first navigation platform
  • Developed a custom route safety scoring engine
  • Successfully integrated real-world OpenStreetMap intelligence
  • Created route comparison based on safety rather than speed alone
  • Implemented emergency SOS escalation workflows
  • Added shake-to-SOS emergency activation
  • Integrated Gemini to provide contextual safety reasoning
  • Created a solution that addresses a real-world personal safety problem

Most importantly, we transformed navigation from a purely efficiency-focused experience into one that actively considers personal safety.

What we learned

Through this project we learned that safety is a complex, multi-dimensional problem that cannot be represented by a single number.

We gained experience working with:

  • Geospatial data
  • Route optimization systems
  • OpenStreetMap ecosystems
  • Real-time mobile application development
  • AI-assisted reasoning systems
  • Human-centered safety design

We also learned the importance of combining deterministic systems with generative AI. Traditional algorithms provide reliability and measurable outcomes, while AI helps communicate complex insights in a more understandable and actionable way.

What's next for AegisPath

Our vision is to evolve AegisPath into a community-powered safety intelligence platform.

Future plans include:

  • Real-time community incident reporting
  • AI-powered hazard detection and classification
  • Dynamic safety heatmaps
  • Crowdsourced safety verification
  • Predictive risk forecasting based on historical trends
  • Live guardian monitoring for trusted contacts
  • Wearable and smart-device integration
  • City-level safety analytics dashboards

We also plan to further expand Gemini integration to provide proactive safety recommendations, adaptive travel guidance, and personalized risk awareness throughout a user's journey.

Our long-term goal is to make safer travel accessible to everyone, ensuring that navigation systems optimize not only for speed, but also for personal security and peace of mind.

Built With

Share this project:

Updates