During a cross country trip, I keyed in a destination that was 2 hours away, through both urban and rural areas and through varying levels of rush hour traffic. I know they have ways of measuring instantaneous traffic levels, but how can they predict road conditions for multi-hour journeys?
In: 8
They collect, analyze, and store all that traffic data to develop forecasts. Not only can they see the current traffic patterns, but they have been storing that data to use to make predictions.
Phones have sensors that can measure speed. You agree to share that data when you use navigation services. Each road has data attached to it describing traffic patterns at different days and times.
When you select a route, the service estimates what the traffic would be like when you hit the road, and adds up all the expected travel times.
They work off past data, combined with whatever available existing data they can get from phones further along the route you’re taking.
When you are using a mapping system real time, if traffic slows, the data returns to the navigation system indicating a slowdown. Think of how many people are using a navigation app, so the data is real time. Many times if traffic comes to a complete stop, before you even get to the slow areas, it may try to re-route you by providing an alternative route to take.
Basically they try their best and then cheat.
They are constantly updating your “time left” with whats going on at the current moment.
So for example, if Google maps says the drive will take 30 minutes the second you leave and you speed the entire time or catch a lucky break in traffic. Let’s say you go twice as fast for the first 10 minutes as you were supposed to.
If that 30 minute thing was just a timer then when you hit start it would just be ticking down 30 minutes. So after 10 it would say 20 minutes left, except it doesn’t. It updates with the new info as it’s happening, so after 10 minutes if you’ve covered 20 minutes worth of distance the “time remaining” on the display will say 10 minutes left.
1 minute left and it says 1 minute, then you get there and it says “arriving now” or “your destination is on the left” or whatever and it seems like it was exactly right the whole time.
You can tell this is what’s happening because sometimes it will think you will be stopped at a red light for 3-4 minutes and you catch a break and it’s green the second you pull up. Watch the timer, it will instantly jump down 3-4 minutes. It’s very effective gaslighting.
The short answer is data… Collecting data on many trips at many times in many circumstances, plus constantly collecting data on current traffic and current events such as road obstructions.
Where you see the accuracy drop is when you drive faster or slower than average. My boyfriend is a bit of an aggressive and zippy driver, so he almost always arrives a few minutes ahead of the estimate.
Side note: This is why Google maps stays so dominant in the realm of mapping and trip planning. The more it builds features and accuracy through data collection, the harder it gets for other app to compete. Every day is another day of data that Google has which a new app wouldn’t. They’ve spent millions of dollars to develop Google maps and I’m actually not sure if they’ve made any profit yet, they definitely ran Google maps at a loss for a long time. But they knew that investing in Google maps early would make it nearly impossible to compete with.
Sorry, I made the short answer long..
Latest Answers