Dawg Path

Inspiration

Walking around campus late at night can be unsettling, especially walking down routes that feel unsafe or are poorly lit. We wanted to make sure students at UW (and beyond) could navigate confidently, not just quickly. Traditional navigation apps focus on the shortest route, but we wanted to prioritize the safest one.

What it does

Dawg Path is a safety-aware navigation app that helps users find routes minimizing exposure to danger zones. It analyzes both road-level danger scores and circle-based safety zones (like well-lit areas or campus patrol zones) to generate paths that balance distance, duration, and safety. Users can toggle between prefer safe and strict safety modes, automatically rerouting around high-risk areas and ensuring they “find their way, the safe way.”

How we built it

We built Dawg Path with a FastAPI backend and Mapbox Directions API integration for real-time route geometry. We used GTFS transit data for Seattle to include campus and city bus routes, and incorporated a custom danger map JSON to weight roads by safety. For enhanced routing, the app fetches multiple walking alternatives, computes a composite safety score from both per-road danger levels and localized safety zones, and ranks the best paths accordingly.

Challenges we ran into

One of our biggest challenges was integrating multiple data sources into a cohesive routing system. Combining Mapbox’s navigation data with GTFS transit schedules and our custom danger zone JSON required careful data normalization and synchronization. We also faced issues with API rate limits and incomplete route geometry, which caused initial crashes during path generation. Another major challenge was calibrating our safety scoring algorithm so that it reflected real-world intuition. For example, making sure a well-lit road wasn’t automatically considered more dangerous than a quiet residential street. Finally, testing and debugging in real time were tough since every change in the scoring logic or zone file could drastically alter routing behavior.

Accomplishments that we're proud of

We’re proud that Dawg Path successfully generates safety-aware routes that adapt to both user preferences and local environmental data. Getting the end-to-end routing pipeline working, from raw GTFS files to Mapbox visualization, was a major technical win. We’re also proud of how cleanly we combined our danger and safety scoring layers into a single model that adjusts walking routes intelligently. Beyond the code, we’re proud of how quickly we learned to think about navigation not just as a matter of where to go, but how safe it feels to get there. Our project feels both practical and meaningful because it addresses a real world concern for student safety.

What we learned

This project taught us how complex real-world navigation truly is. Balancing speed, accuracy, and safety required us to think beyond traditional pathfinding. We learned how to merge multiple geospatial data formats, use APIs efficiently under time constraints, and quantify an understanding of “safety” using various important factors. On the teamwork side, we learned how to divide and conquer effectively, with one member focused on data ingestion, another on the backend safety logic, and another on visualization. Most importantly, we realized how technology can meaningfully enhance personal security and how thoughtful design choices can make digital navigation more human-centered.

What's next for Dawg Path

Looking ahead, we plan to make Dawg Path even smarter and more community-driven. Our next steps include partnering with local police departments to integrate live safety and incident data, ensuring that our route recommendations reflect real-time conditions. We also want to increase crowdsource safety insights, allowing users to contribute reports or feedback that improve future route scoring. By combining official data with community input, Dawg Path can become a dynamic reflection of how people actually experience their environment. Finally, we’re exploring geofencing and traffic-aware features to show congestion, blocked routes, or high-activity areas in real time, giving users even more control over how they navigate safely through their city.