Resilience and rewiring of the passenger airline networks in the United States

arxiv.org

23 points·luu··4 comments

That's a lot of work to come to the conclusion that a mostly point-to-point network is more resilient to attack/failure than a hub-and-spoke network. Seems obvious that losing a hub creates a mess. Am I missing some nuance here?

I think the big thing is that hub and spoke was always considered more efficient but here they are saying that if you take into account resilience into the equation then p2p may be more efficient than a hub and spoke model.

I think point-to-point is supposed to be the most efficient provided you accurately match airplanes to the routes people want to fly.

Hub flying is useful for folks that want to go from one small city to another; on a given day there might be two people who want to fly from Providence, RI to San Antonio, TX so it's not viable to fly a plane between them, but if you route them via Philadelphia, PVD-PHL-SAT, you can accommodate them with limited extra distance or extra time and also accommodate people wanting to travel PVD-PHL and PVD-PHL-SYR, etc.

Once you start working in units of "planeloads" instead of "people", point-to-point flying is likely going to be more efficient in terms of time, distance, cost, etc...

This is actually quite significant if you consider the perturbation which weather systems can have on flight operations. Weather can easily lead to annual losses > 100m to individual US carriers, due to delay and cancellations.

Reply on news.ycombinator.com