By: John Croft | Washington | Aviation Week | May 1-14, 2017:
With a blizzard set to slam the Northeast U.S. on a Monday night in mid-March, airlines on the preceding Saturday began pre-emptively canceling flights.
Flight-tracking website Flightaware.com listed more than 9,000 operations that had been canceled between Sunday and Wednesday, when the recovery began. American Airlines ultimately proactively canceled 1,990 mainline and regional flights on Tuesday and 519 on Wednesday of that week.
For American, what might seem like a brash decision—canceling flights based on forecasts several days into the future—was in fact based on a risk assessment steeped in data and simulation rather than the ad hoc, experience-driven response of the past.
“That’s the kind of process we’ve now moved into for managing through these winter storm events,” says Tim Niznik, American’s director of opera¬tions systems and decision support.
“We’re converting the forecast into something we can simulate—and do what-ifs against—much earlier in the process to hopefully make sure we’re well positioned with the right number of operations with the right kind of deicing personnel on hand to manage through that event.”
The data-driven approach to de-icing is part of a broader push within the airline to better manage disruptions, eventually with the help of probabilistic forecasts and big data.
To read full article – please click here.
