Getting Aviation On the Bus
Real-time intermodal substitution (RTIMS) as an airport congestion management strategy: putting passengers on deluxe motorcoaches in periods of airport shutdown.
The massive weather-related disruptions this winter that led to profuse apologies (and refund offers) from airline CEOs, including low-cost carrier JetBlue, and calls for congress to pass a passenger bill of rights might have been avoided. All the airlines had to do was cancel some flights into the socked-in airports and put their passengers on a bus.
That's the conclusion reached by researchers at ITS Berkeley's NEXTOR aviation unit, the National Center of Excellence for Aviation Operations Research.
The scheme, known as "real-time intermodal substitution" (RTIMS), was developed by NEXTOR PhD student Yu Zhang with NEXTOR Co-Director Mark Hansen. It is part of Zhang's PhD dissertation.
Under Federal Aviation Administration (FAA) policy, airlines in this age of de-regulation are encouraged to respond to disruptions in the National Airspace (NAS) on their own before the FAA steps in. One response airlines can choose is to cancel flights.
By using RTIMS, Zhang and Hansen explain in their paper delivered to the 2007 meeting of the Transportation Research Board, airlines might find cancellations more affordable and more productive, both by cutting passengers' total travel times, and by preventing hub airports from becoming too congested with incoming planes that can't fly out.
The challenge is to calculate which flights to cancel and the best alternate transportation for the people flying on them. That is where the buses, or luxury motor coaches, as Hansen explained, come in. Short-haul commuter flights would be canceled and their passengers put on motor coaches operated under the oversight of the airline. The goal is to achieve the "optimal" solution, that is, reducing delays to the largest degree at the least "cost" to airlines, airport operations and passengers. (For details on the model and the cost allocations, download the paper or view a pdf of the PowerPoint presentation.)
A look at the numbers
Consistently over the years, some 70 percent of the total delays in the NAS are due to weather. And a small fraction of the total number of passengers whose itineraries are disrupted suffer disproportionately from weather delays. Three percent of the total passengers have their travel disrupted by weather, but they suffer 39 percent of the total delay, with an average of 303 minutes, versus 16 minutes for the passengers whose trips are not disrupted.
The solution is simple: when weather causes traffic to slow at hub airports, canceling the short-haul flights to those airports and putting the passengers on buses gives airlines more flexibility.
Also, Zhang explained, "By focusing on short-haul flights, we are working with the flights that have the greatest ability to adapt to real-time information about weather and the least advanced planning necessary." In some case, even those who end up traveling by bus will end up arriving sooner than if they'd stayed on the plane waiting to take off or while it was diverted.
"There is extra travel time because they are taking the slower mode, but because the slower mode is not subjected to the capacity constraints, everyone is better off," she explained.
Sometimes, of course, bus riders will experience a longer delay than if they'd stayed on the plane, but the rest of the system will benefit.
"We want to use this idea to improve the stability of the transportation system," Zhang said, and reduce the vulnerability incurred by its reliance on a single mode, as was the case after the Sept. 11 attacks when the entire air system was shut down.
Zhang proceeded to "solve" the problem of how to assign flights to be canceled in order to reach the optimal level of performance balanced with costs.
Airlines have a financial incentive to avoid delays and unnecessary cancellations, beyond the goal of keeping passengers happy. Although the figures are subject to fluctuation, delaying a flight can cost on the order of $500 an hour. Canceling a flight incurs costs upwards of $10,000 for a medium-sized 159-seat plane. Short-haul flights cost less to cancel because less repositioning is required, and fewer passengers need to be re-accommodated. Deploying a motor coach costs about $100 an hour.
The RTIMS approach focuses on the closest airports that are feeding the hub. The airline that canceled the flight would also operate the motor coaches, so it would not be losing passengers to other airlines, under this scenario.
The choices in the model are: Cancel or fly it? If cancel, is the flight substituted with surface transportation, or do the passengers take a subsequent flight?
Testing the Theory at SFO
Zhang ran her proposal for a typical day at SFO with a feeder airline operating the short-haul segments into the airport, for a total of about 130 flights, when the airline comes close to shutting down because of morning fog.
Three scenarios are compared: one is the "doing-nothing" scenario, that is, no flight cancellations, the second is optimizing flight cancellations by minimizing total disruption costs, but not using alternative ground transportation, and the third one is RTIMS, which is very similar to the second one, but uses ground transportation modes as alternatives. With RTIMS, disruption costs for the airlines are 50 percent lower than with the first "doing-nothing" scenario. With the same optimization methodology, RTIMS shows savings of between 8 and 15 percent compared to the second scenario.
The researchers conclude that this approach has "significant promise."
Another application would be to combine the intermodal approach with a reliever hub network, where airports in the vicinity of the hub can be used to land planes and connect passengers to their destinations through surface transportation links.