What causes flight delays? The reason could be anything from extreme weather to air traffic congestion to aircraft maintenance issues or something else entirely. The bottom line is, flight delays are extremely frustrating for weary travelers stuck at airports. So, Google is updating its Flights service with an awesome new feature – using machine learning algorithms to predict delays in departure times.

The Flights app is normally populated with the information Google pulls in from the airlines directly. But for this feature, Google has turned to historical flight status data and artificial intelligence technologies to predict when a flight will be delayed.

The algos will comb the data to see what parameters are common between delayed flights – location, weather conditions, aircraft arriving late from the previous destination, etc. Once Google is at least 80% sure that these factors are conspiring in a manner that an airplane will be delayed, it will flag that information to the Flights app, and also specify the likely reason for the holdup – without waiting for or depending on the information coming in from the airlines.

To be on the safe side, Google still recommends getting to the airport with enough time to spare.  But it hopes to give travelers enough information to manage expectations and prevent any unpleasant surprises.

Now, Google is not the first or only one trying to make the travel sector less complex with machine learning. Back in 2013, researchers at Singapore’s Institute for Infocomm Research used machine learning techniques to beat the industry benchmark for the estimation of the arrival time of domestic flights by 40%. Last year, Eastern Macedonia and Thrace Institute of Technology in Greece showcased how it was feasible to predict prices for flights based on historical fare data. But Google’s application of machine learning is certainly taking the technology mainstream like none other.