WayGuard

Inspiration

WayGuard started from a simple question: what if navigation could optimize for personal safety context, not just time and distance?

Most map apps treat all streets as equal, but city conditions vary by place and time. We wanted to build a tool that helps people make more informed route choices using open civic data, especially for walking and late-night travel.

What it does

WayGuard is a location-aware navigation and analytics platform that recommends multiple route options and scores them using a blend of travel time, distance, and risk exposure. Users can tune a risk-aversion control from fastest to lowest-risk behavior, choose which datasets to include (NYPD, 311, and restaurant inspections), and adjust relative weights. The app presents these options on an interactive map with clear route metrics and explanation UI, and it also includes an analytics view with trend visualization and a city heatmap so users can explore broader incident patterns.

How we built it

We built WayGuard with a FastAPI backend and a React + Vite + TypeScript frontend. On the backend, we reused and extended our Python geospatial/data logic, added route scoring over an H3-based risk surface, and exposed APIs for autocomplete, navigation computation, and analytics summaries. On the frontend, we implemented a planner-first navigation workflow, map-centered route visualization, route comparison tables, and a polished design system with modern components, loading states, and responsive behavior.

Challenges we ran into

One of the biggest challenges was handling real-world civic data quality and timing differences across sources. Datasets update on different cadences, can be sparse in different regions, and can produce misleading outputs if date windows aren’t aligned. We also had to harden around routing provider constraints, edge-case failures, and early scoring issues where risk values could flatten unexpectedly. On top of that, shaping a complex set of controls into a UI that felt clean and intuitive required several iterations and close attention to interaction design.

Accomplishments that we're proud of

We’re proud that WayGuard grew from a prototype into a polished full-stack product. We built a route system that clearly balances time, distance, and incident-density risk, while giving users transparent controls for risk aversion, dataset selection, and weights. We also unified multiple civic datasets into one explainable experience and delivered a modern, map-first UI that makes complex geospatial insights feel intuitive and usable.

What we learned

This project taught us that building trustworthy geospatial products is as much about communication and UX as it is about modeling. Safety-oriented recommendations must be explainable, ethically framed, and resilient to imperfect data. We also learned the value of modular system design: separating backend logic from frontend interaction made it much easier to iterate rapidly on both product quality and technical reliability.

What's next for WayGuard

Next for WayGuard, we want to deepen personalization and realism by adding time-of-day/day-of-week risk modeling, richer explainability around why a specific route is recommended, stronger confidence indicators by area, and more robust performance through smarter caching and incremental refreshes. We also want to validate product impact with user testing and prepare the platform for a broader pilot deployment with real feedback loops.