AI-Optimized Public Transport Network
Fixed schedules meant buses were stuck in traffic or running empty. Commuters abandoned the service for taxis.
Buses were running empty on some routes and overcrowded on others. The static schedule didn't match the shifting pulse of the city.
•City-wide fleet
•Public budget
•Citizen satisfaction
•Carbon reduction
Demand Modeler
Predicting ridership using ticket swipes, weather, and mobile data.
Dynamic Scheduler
Adjusting bus frequency every hour to match predicted crowds.
Citizen App
Live ETA and crowding indicators for passengers.
1
Union Rules
Schedules had to respect driver break times and shift lengths.
2
Legacy
Integrating with 10-year-old GPS trackers on buses.
15%
increase in total ridership
10%
reduction in fuel costs/emissions
Improved
on-time performance by 20%
Higher
citizen satisfaction scores
We shifted to a Dynamic Schedule. AI predicts where people need to go. We add buses to busy lines and cut empty ones. It makes public transport faster and more reliable than driving.
PythonPandasGoogle OR-ToolsReact NativePostgreSQL
