Wander Lens AI - Smart Route Planning with POI Discovery

Experience your journey before you go - Plan routes, discover places, and preview scenery along the way

Inspiration

The inspiration for Wander Lens AI came from a team member's personal experience taking a trip between Los Angeles and San Francisco. They discovered there's a big difference between taking the scenic coastal routes (Highway 1 and 101) with breathtaking ocean views, charming coastal towns, and stunning vistas versus the inland Highway 5 surrounded by farmland and, well, cow dung.

We decided it would be perfect to experience the trip before actually going - to see what restaurants we'd pass, where we could stop for coffee, what attractions were along the way, and most importantly, preview the actual scenery we'd encounter. Traditional GPS apps tell you how to get there, but they don't help you experience the journey or discover what makes each route special.

Wander Lens AI was born from this simple question: "What if you could preview and plan every aspect of your road trip, from the route itself to every interesting stop along the way?"

What it does

Wander Lens AI is an intelligent route planning application that goes beyond simple navigation:

🗺️ Smart Route Planning

Multi-waypoint routing with automatic segment merging for complex journeys
Multiple routing profiles: Trekking, Fast Bike, Safety, Shortest, Moped
Real-time route metrics: Distance, estimated time, elevation gain/loss
Interactive map interface: Click to add waypoints, drag to adjust

🔍 POI Discovery

Discover interesting places along your route with smart filtering:

🍽️ Restaurants - Find dining options along the way
☕ Cafes - Perfect spots for coffee breaks
🍺 Bars & Pubs - Evening entertainment stops
🅿️ Parking - Convenient parking locations
🚻 Restrooms - Essential facilities
🏛️ Tourism & Attractions - Museums, artwork, landmarks
➕ And More

📸 Scenic Preview

Street-level imagery from Google Maps to preview your route's scenery
See what the drive actually looks like before you go
Discover scenic viewpoints and photo opportunities

🎯 Smart Filtering

Filter POIs by type to find exactly what you need
Real-time updates showing available options
Click any POI to focus and get details

How we built it

Backend Architecture

Built with Python/Flask, our backend uses three powerful APIs:

1. BRouter API - Route Generation

Generates optimized routes with multiple profiles
Handles complex multi-waypoint journeys
Implemented intelligent segment splitting/merging for routes exceeding API limits
Returns detailed GeoJSON with elevation and timing data

2. Overpass API - POI Discovery

Queries OpenStreetMap for points of interest
Searches within route bounding boxes
Filters by amenity types (restaurants, cafes, parking, etc.)

3. GoogleMaps API - Street Imagery

Provides street-level photography along routes
Offers panoramic and directional views
Helps users preview scenic sections

Frontend

React with TypeScript for type-safe component development
Leaflet for interactive mapping and visualization
Vite for fast development and optimized builds
Responsive design optimized for desktop trip planning
Lovable for site design

API Design

RESTful endpoints for clean integration:

GET /api/route?waypoints=lon1,lat1;lon2,lat2&profile=trekking
GET /api/pois?bbox=minLat,minLon,maxLat,maxLon
GET /image-data?lat=37.7749&lon=-122.4194

Challenges we ran into

1. Multi-Waypoint Route Complexity

BRouter's public API has limitations on complex multi-waypoint requests. We solved this by:

Automatically detecting when routes exceed limits
Splitting into individual segments
Intelligently merging coordinates without duplication
Aggregating metrics (distance, time, elevation)

2. POI Data Overload

Overpass API can return thousands of POIs for large areas. Our solutions:

Smart bounding box calculation from route geometry
Client-side filtering for real-time responsiveness
Prioritized display showing most relevant results first
Implemented pagination for large result sets

3. API Rate Limits

Managing multiple external APIs required:

Intelligent caching strategies
Request batching and throttling
Graceful error handling and user feedback
Fallback mechanisms for service interruptions

4. Coordinate System Complexity

Different APIs use different coordinate systems:

BRouter uses lon,lat format
Leaflet uses lat,lon format
Overpass uses lat,lon for bounding boxes
Implemented consistent transformation layer

Accomplishments that we're proud of

✅ Successfully integrated three complex APIs (BRouter, Overpass, KartaView) into a cohesive system

✅ Built multi-waypoint handling that automatically splits and merges routes, solving a major limitation of the BRouter public API

✅ Created an intuitive UI where users can plan routes and discover places with just a few clicks

✅ Implemented smart filtering allowing users to customize their trip based on their interests

✅ Clean, maintainable codebase with separate service modules for easy future expansion

What we learned

Technical Insights

API integration requires careful abstraction - Each service has unique quirks; wrapping them in clean interfaces made development much smoother
Geographic data is complex - Coordinate systems, projections, and bounding boxes require careful handling
Real-time filtering > Server-side filtering - For responsive UIs, client-side filtering of moderate datasets provides better UX
Error handling is crucial - External APIs fail; graceful degradation and clear error messages are essential

Product Insights

Users want to experience routes, not just follow them - The combination of scenic previews and POI discovery addresses a real gap in existing navigation apps
Filtering is essential - Not everyone wants to see every restaurant; letting users choose what's relevant makes the app much more useful
Interactive maps are intuitive - Click-to-add waypoints feels natural and reduces friction

Development Process

Start with working examples - Building simple demos first helped us understand API limitations before tackling complex features
Test with real data - Using actual SF coordinates revealed issues we wouldn't have found with synthetic data, and was much quicker
Incremental integration - Building one service at a time and validating before moving to the next prevented compounding errors

What's next for Wander Lens AI

Short Term (MVP+)

🎨 Enhanced UI/UX - Polish the interface with better visual design and animations
📱 Mobile Responsiveness - Optimize for mobile trip planning on-the-go
💾 Trip Saving - Allow users to save and share their planned routes
🗺️ Route Alternatives - Show multiple route options with pros/cons

Medium Term

🏨 Accommodation Integration - Find hotels/camping along routes
⛽ Gas Station Locator - Essential for long road trips
🌤️ Weather Integration - Check weather conditions along the route
📊 Route Comparison - Side-by-side comparison of different routes
⭐ POI Ratings - Show reviews and ratings from various sources

Long Term

🤝 Social Features - Share trips, follow other travelers' routes
🎯 AI-Powered Recommendations - ML-based suggestion of stops based on preferences
🚗 Multi-Modal Transport - Support for mixed transportation (drive + bike + walk)
📱 Native Mobile Apps - iOS and Android apps with offline capability
🔌 EV Charging Stations - Essential for electric vehicle trip planning
🎵 Spotify Integration - Road trip playlists synced to journey segments