Attention Marketers: Access 30 Million IT Decision Makers with ECT News Network's INSTA-LEADS ∑ Click to Learn More!
Welcome Guest | Sign In

FlightPredictor: Don't Leave Home Without It

FlightPredictor: Don't Leave Home Without It

Ever get the feeling an airline is placating you with optimistic predictions of flight statuses that more often than not turn out to be wrong? You don't have to be at the mercy of the airlines any longer. With FlightPredictor, you can draw upon a wealth of data fed into a killer algorithm that results in flight status information that actually comes close to reality.

By Patrick Nelson LinuxInsider ECT News Network
02/01/13 5:00 AM PT

FlightPredictor, a mobile app from Gregstoll, is available for US$1.99 at Google Play.

FlightPredictor Gregstoll's FlightPredictor app takes FlightCaster.com's powerful flight prediction data and wraps it into a mobile app, in this case, for Android -- and what a great job Gregstoll, FlightCaster and its data supplier FlightStats have done.

FlightPredictor for Android uses historical flight data going back 10 years, along with real-time conditions like in-bound airplane tracks, weather, and live information from FlightStats and the Federal Aviation Authority to predict flight delays.

Predictions are based on learning techniques and are generated by an algorithm. That data is then delivered to the app.

How It Differs

One of the ways FlightPredictor differs from the sometimes dubious airline-provided information is that it relies heavily on delay factors, rather than simply whether there's an aircraft at the gate or not.

Plus, its independence from the airline means you're not subject to the airline's ulterior motive -- that it may want you hanging around the gate so it can board you quickly, for example.

Algorithm Factors

Delay factors include inbound aircraft, departure airport status, arrival airport status, official airline status, departure weather and arrival weather.

Sources include Bureau of Transportation Statistics, FAA Air Traffic Control System Command Center, FlightStats and the National Weather Service.

Based on all of these elements, including historical data, the algorithm can provide a probability of delay based on a percentage.

For example, it can indicate that the predicted arrival status of a particular flight will be probably delayed with percentages as follows: 3 percent probability of it being on time; 14 percent probability that it will be less than an hour late; and 83 percent that it will be more than an hour late.

In other words, even if the airline reckons you're going to get to where you're going on time -- you ain't gonna.

In Use

I used the app on a recent flight from New York's Kennedy into Los Angeles and was impressed by the flexibility of the app.

Early in the day, the app can show a flight as having, say, an 86 percent chance of being on time even though the airline hasn't posted the inbound aircraft. The app just uses whatever data is available to make the prediction. As more data comes online, it throws numbers into the equation, providing more accurate predictions.

What's really interesting is that the app and its data focus on real life -- not an airline's delusional prediction of real life. For example, FlightPredictor ignores an airline's Walter Mitty-like fantasist's assumption that once you're on the plane and the doors are shut, you're going to take off and be on time. We all know that is often not the case due to ground delays.

I can't count the number of times I've heard the cockpit optimistically announce an on-time arrival subject to air traffic delays -- only for there to indeed be air traffic delays. What a surprise.

User Interface

The app screen is messy and could use a UI cleanup. However, it does contain massive amounts of at-a-glance data: flight number; departure gate; arrival gate; predicted arrival status; official airline-reported status of estimated departure and estimated arrival -- including the number of minutes that the airline estimates the flight to be late; and delay factors like inbound aircraft, FAA alerts and weather for the departure and arrival airport -- and so on. Great Stuff.

Bonus features include 30 airport maps and a sharing function where you can send flight prediction reports by email -- to a family member, say.

Unfortunately, the app only works with U.S. domestic flights.

In Conclusion

Over the years I've spent countless hours -- probably adding up to days out of my life -- hanging around airport departure screens, being thrown the occasional tidbit of information from arrogant airlines. It's only been in the last few years that live gate and boarding information have been available to mobile users.

Throw apps like FlightPredictor into the mix, and everything changes. Power to the People!


Patrick Nelson has been a professional writer since 1992. He was editor and publisher of the music industry trade publication Producer Report and has written for a number of technology blogs. Nelson studied design at Hornsey Art School and wrote the cult-classic novel Sprawlism. His introduction to technology was as a nomadic talent scout in the eighties, where regular scrabbling around under hotel room beds was necessary to connect modems with alligator clips to hotel telephone wiring to get a fax out. He tasted down and dirty technology, and never looked back.


Facebook Twitter LinkedIn Google+ RSS