College project, Python

Route
Master

Vehicles finding their way across a real road network, with nothing but the standard library behind it.

157 nodes373 edges69 expanded22 hops
Search
A* with a straight-line heuristic
Neighbours
Spatial hash written from scratch
Data
OpenStreetMap road geometry
Dependencies
None. Standard library only
github.com/Brkic365/routeMaster
The RouteMaster simulation dashboard

Built for college, in Python, with a rule I set myself: no external libraries. Everything runs on the standard library, including the pathfinding, the spatial indexing and the drawing. That is also why it looks the way it does. Tkinter is not going to win any design awards, and I decided early that I would rather have the algorithms be the interesting part than spend the time making it pretty.

Pathfinding across a city is cheap the first time. It gets expensive when hundreds of vehicles each need a fresh route every time something changes, and the plain neighbour search starts eating the whole frame. Bucketing the network into a spatial hash turned that scan into a lookup, which is what made it possible to run the simulation and watch it at the same time.