If you have any kind of technical ear to the ground these days, you won't go too far without bumping into the term "Big Data". Originally spawned from web giants like Google, Yahoo, Amazon and Facebook, it was a term coined for how they can analyze and look at all the user information they gather as you use their systems to search, shop, email and share stuff online with your friends.
They had to come up with a new way to store and search through all this information as traditional relational databases couldn't handle the volume and types of data generated on these sites.
Why do they want to go through all this information? Well, mostly to be able to sell advertising and/or additional products to you by understanding more about you. This "prediction" of what you might like is the other end of the Big Data picture, called Predictive Analytics.
While that is all great for advertising, what can we do with Big Data and Predictive Analytics in the air transport industry? To re-use a well-worn phrase, "the sky is the limit", as one starts to put some imagination to this question. There is certainly a lot of data floating around the industry.
There's the obvious passenger information like reservations and frequent flyer records, which many airlines have been doing traditional "data mining" on for some time. There is all the operational information, upwards of 70,000 commercial flights per day. As well as not so obvious information, like the flow of smartphone WiFi radio signatures moving through an airport as passengers arrive and depart (see my previous post on geolocation).
From all this information, both historical and the real-time streams of data, what can we predict about the future? Fortunately, the phrase "history tends to repeat itself" is, more often than not, true - and the answer is "A LOT"!
Certainly, predicting potential "disruptions" that could cause ripples in the normal operational flows of flights and people can be used to take action to prevent or at least better prepare for what is about to happen. The earlier the prediction can be made and the higher the accuracy, the more impact it can have in allowing adjustments that minimize the impact, both at a customer-experience level and financially for the airport and airlines.
This is an area that SITA Lab will do some work on in 2013. So, as always, watch this space, and for a more humorous explanation of predictive analytics, I pass you to my friend Stephen Colbert.