• Vehicles 0
• On board 0
• Waiting 0
• Served 0
• Rejected 0
• Mean wait 0
• Mean occupancy 0

Mobility-on-demand simulation on the streets of central Prague: random Poisson demand, dispatched by an insertion heuristic, vehicles at 8 m/s; the number beside a vehicle is the passengers on board, the speed slider runs the clock faster than real time. Runs entirely in your browser; changing the fleet size restarts the run. Map data © OpenStreetMap contributors.

• Stations 0
• Vehicles 0
• Requests 0
• Deadheading 0
• Per trip 0
• MILP …
• Heuristic …
• Solve time

Station positioning: every idle vehicle waits at a station and returns to one after each trip, so the stations should sit close to where trips start (orange) and end (blue). The squares are the chosen stations with their vehicle count; thin lines show which station serves each point. Solved as a p-median MILP with javascript-lp-solver in your browser, next to a greedy-plus-interchange heuristic. Click the map twice to add a request of your own: first the origin, then the destination.