Google Flights can predict flight delays and warn you BEFORE your airline
Using machine learning algorithms, the search engine giant has devised a way to guess when your flight will be delayed – and can even tell you why
GOOGLE can now use machine learning to guess if your flight will be delayed or cancelled, with startlingly high accuracy.
A new update to the Google Flights tool can warn you ahead of time, even before your airline has told notified you about delays.
The update, which is now live to users, checks historical flight data to work out if your journey is at risk.
"Around this time of year, snowbirds fly out on tropical getaways, skiers and snowboarders take a weekend to hit the slopes, and potential flight delays are stressful for all travellers," explains Anket Mathur, Flights Product Manager at Google.
"Google Flights can now help you predict those flight delays."
Google basically combs through loads of historical data to find common patterns or themes that lead to delays.
It will use info based on locations, the weather, and aircraft arriving late to the airport, so it can more accurately predict when your flight will be delayed – and by how much.
To make sure it doesn't mislead you, Google will only warn you when its algorithms are "80% sure there'll be a delay".
It says that the algorithms are so advanced, they can "predict some delays even when this information isn't available from airlines yet".
"We still recommend getting to the airport with enough time to spare, but hope this information can manage expectations and prevent surprises,"; Mathur adds.
MOST READ IN TECH
To access Google Flights, you simply need to search for your flight – or airline and flight route – on Google, and the details will appear right in your search results.
If it doesn't show up, you can also click the 'Flights' option at the top of Google's search page, which lives alongside other services like 'News' and 'Images'.
It's worth noting that Flights already warned about delays, but the new addition involves predicting the delays – and showing the reasons behind those predictions, too.