Strava Route Xplore

Strava Route Xplore reveals where people actually run and ride by turning thousands of activities into a live, interactive map of route popularity—so you can discover safer, smarter paths instantly.

Comment

Inspiration

Strava makes it easy to log activities, but hard to understand where people actually go at a route level. I wanted a way to see real-world usage patterns — not just individual runs or rides — so people could pick safer, more popular, and more interesting routes.
This started after repeatedly choosing empty or unsafe cycling paths just because I had no visibility into which routes were truly well-traveled.

What it does

Strava Route Xplore visualizes aggregated route usage from your Strava activities.

It:

  • Decodes and aggregates activity polylines into shared route segments
  • Colors each segment based on how often it's traveled
  • Lets you click any point on the map to see which activities passed through it
  • Supports filtering by sport, year, and other metadata

The goal: turn individual activity noise into clear route intelligence.

How we built it

Frontend

  • React + Vite + TypeScript
  • react-leaflet for interactive maps
  • polyline for decoding Strava GPS data
  • Client-side aggregation + caching using localStorage

Backend

  • Python + FastAPI + Uvicorn
  • Handles Strava OAuth and token refresh
  • Provides endpoints for activities + auth flow

APIs

  • Strava API for activities + polylines
  • OpenStreetMap for map tiles

Key design choice: heavy processing happens on the client to reduce backend complexity and keep the app responsive.

Challenges we ran into

  • Stack overflows from recursive data processing on large datasets → refactored to an iterative approach.
  • GPS noise & floating point issues → solved using coordinate quantization to merge near-identical segments.
  • Strava rate limits + expiring tokens → built a refresh-token backend flow with auto-retry logic on the frontend.
  • Rendering thousands of segments → optimized via aggregation and plan Web Worker + Canvas rendering for scale.
  • Map re-render and fit loops → fixed with memoization + render guards.

Accomplishments that we're proud of

  • Built a full end-to-end working system: auth → data → visualization.
  • Implemented click-anywhere → see activities (our main “wow” feature).
  • Designed a robust segment aggregation system instead of just drawing raw tracks.
  • Handled real API and geospatial scaling challenges under time pressure.

What we learned

  • Recursive logic doesn’t scale well in geospatial datasets — iterative + indexed solutions are more stable.
  • Visual clarity matters just as much as algorithm correctness.
  • OAuth flows break easily — refresh logic must be designed early, not patched later.
  • Small interactions (like clicking a point to see activities) massively increase perceived value.

What's next for Strava Xplore

  • Add Web Worker + Canvas renderer for large-scale performance.
  • Introduce a heat legend and user-tunable smoothing options.
  • Filters by time-of-day, day-of-week, and activity type.
  • Shareable route visualizations and GPX export.
  • Potential premium features for clubs, race organizers, and urban planners using route density data.

Long-term vision:
Become the “route intelligence layer” on top of activity platforms like Strava.

Share this project:

Updates