Researchers have built a demand-driven model for siting the vertiports that would anchor an urban air mobility network - a fancier term for flying-taxi infrastructure.
The study estimates travel demand from commuter and passenger activity data, clusters likely trip start and end points with K-means to find candidate vertiport sites, then filters those sites by eVTOL range and minimum spacing between stations. A discrete-event simulation then tests each network design, modeling vehicle dispatch, battery swaps, and deadhead flights - the empty repositioning trips vehicles make between stations. In a Greater Los Angeles case study, the best-performing network grows from four stations and four aircraft at low demand to sixteen stations and twelve aircraft at the highest demand level tested.
More stations and aircraft made service faster and more regular, but did not eliminate deadhead flights - the model shows spatial mismatches between where trips start and end keep forcing empty repositioning runs, an ongoing operational cost. The researchers also found flying only saves meaningful time over ground travel on longer or already-congested routes, once you subtract time spent getting to and from the vertiport.
That's a narrower use case than the industry's demo videos suggest, and it comes from a simulation, not a fleet in the sky - so treat these numbers as a planning exercise, not a launch date.