PennyFarething

Inspiration

Commuting in London often forces a trade-off: speed, cost, or comfort. At the same time, many short journeys that could be cycled end up being fully spent on crowded trains. We were inspired by the idea that small changes - like cycling just part of your journey - could make commuting not only faster and cheaper, but also healthier. We wanted to make those hybrid journeys easy to discover.

What it does

PennyFarething finds the best routes across London by combining cycling with public transport. Instead of defaulting to the nearest station, it explores all stations within cycling distance of your start and end points.

It then surfaces the best options by balancing:

⏱️ Total journey time
💷 Cost
🚴 Cycling effort

The result is a set of optimised routes - including the fastest, cheapest, and best value - that often beat traditional transit-only journeys while helping users stay active.

How we built it

We built a full-stack web app using:

Next.js for the frontend and API routes
Google Maps APIs (Distance Matrix, Directions, Places) for cycling times, routing, and geocoding
TfL APIs for live transit journeys, stations, and fare data

At the core is a routing engine that:

Finds nearby stations using the TfL StopPoint API
Batch-computes cycling times using Google Distance Matrix
Filters viable station pairs
Fetches transit routes and fares in parallel
Applies a Pareto frontier algorithm to return only the best trade-offs between time and cost

We only fetch detailed map polylines for the final routes to keep things fast and efficient.

Challenges we ran into

API coordination: Combining multiple APIs with different formats, limits, and response times required careful orchestration
Performance: The number of possible station combinations grows quickly, so we had to aggressively batch, prune, and parallelise requests
Fare accuracy: Getting reliable fare estimates from TfL and aligning them with journey results was tricky
Balancing UX vs complexity: Presenting multi-modal routes in a simple, intuitive way without overwhelming the user

Accomplishments that we're proud of

🚀 Building a working end-to-end routing engine that genuinely improves real journeys
🧠 Implementing a Pareto optimisation system to intelligently filter results
⚡ Reducing API calls with batching (cycling times in just two requests)
🗺️ Clean, interactive UI that clearly visualises hybrid routes
💡 Creating a product that encourages healthier, more active commuting

What we learned

How to design and implement multi-objective optimisation systems in a real-world context
The importance of API efficiency and batching at scale
How small UX decisions can make complex data feel simple
That combining datasets (cycling + transit + fares) unlocks insights users wouldn’t find otherwise