Google Maps updates itself using AI | Tech for Product Managers
3 key Concepts
The world keeps on changing at a very rapid pace.
And Google Maps has to represent this ever changing world in a very easy and intuitive way.
Have you ever wondered how Google Maps keeps evolving as the world changes?
In this article, we will learn about the following things
How Google Maps maps the Entire World, the role of Machine Learning in mapping?
How AI and Images help Google Maps in self-updating
How AI helps predict traffic and Routes
How Google Maps maps the Entire World?
There are several ways in which Google Maps takes the input data and maps the world.
The various contributors are — Satellite Images, Street View and Human uploaded Data ( this includes both the Data Operation team and the contributions by the different people )
Satellite Images
Satellite Images give a very good picture of the Roads, Buildings and different regions. This may not be very clear, but it gives a very good sense of how the landscape looks.
Street Images
Another one is the Street View and the Images — these are contributed by the different sources as in Data operation team, the Car mounted with the camera, Bag backpack trackers who mount the camera with them to show you the heights. In the case of the dessert area, the cameras are mounted on the Camels. These Images help a lot in mapping the entire Google Maps. Google Maps has created 170 billion Images till now
Contribution by People
Finally, there is the contribution which we make while using Google Maps. Google Maps receives around 20 million contributions each day from people using Google Maps.
Machine Learning helps Google Maps a lot in mapping the entire world.
Role of Machine Learning
So the satellite Images give a broad sense of the properties like buildings, parks, layout, etc correct?
But there are billions of such properties.
How do Google Maps know which one is a building and which one is a park?
And the building is more than just a landmark, people associate it with proper shape and edges.
Earlier algorithms used to just fuzzy mark, which is one of the properties is building, but now they are using a machine learning model to give a proper shape and size to it as well. This is depicted in the diagram.
How AI helps Google Maps in self-updating
Think about how many things change around us all the time.
New businesses open, old ones close, and even roads can get a facelift. Keeping a map accurate with all that is happening is a huge job! That’s where AI and images come to the rescue.
One really neat way Google Maps stays current is by automatically checking and updating business hours.
Imagine a little lemonade stand, let’s call it ABC Lemonade Shop. Now, things like pandemic restrictions can make business hours change a lot. Google Maps uses a smart computer system (that’s AI!) to figure out if ABC’s hours listed on the map might be wrong.
How does it do that? Well, it looks at a few things
When was ABC last updated with its business info?
What time are other lemonade shops nearby usually open?
The Popular Times for ABC’s shop — you know, those bars that show when it’s usually busy. If it says it’s closed during its usual busy time on Thursday afternoon, something’s probably not right.
Then, the AI gets to work like a detective!
It looks at the hours of other nearby lemonade places, checks out ABC’s website if it has one, and even scans Street View images of his shop looking for any signs with the business hours on them.
How is Google Maps able to predict Traffic?
One of the biggest use cases, which Google Maps solves, is to help users navigate from one place to another.
When people move from one place to another. Aggregate data can be used to get the traffic conditions, but it will not tell you what will happen after 20 minutes or 30 minutes in the drive correct?
So, Here Google uses machine learning to understand current traffic conditions and historical traffic patterns.
This historical pattern can change with time as well.
The traffic conditions of a certain place change concerning time. So Google Maps combines both the information to predict the future traffic condition, and with the help of this, Google Maps can give a certain ETA.
But any anomalies require changes in the model. For example, during the lockdown due the COVID, there was an external factor that changed the course of traffic for a longer time. So Google Maps updated its model.
For Such Case Studies — Download our Book ( Tech for Product Managers
Resources to Download
Strategy for Product Managers, Leaders, Everyone ( Download Here )
PM Interview Mastery Course ( Crack Interview like top 1% ) — First Principle Thinking ( 10 Video lectures )
Technomanagers
More about PM Interview questions and Mock Interviews | YouTube | Website | Tech & Strategy Newsletter
To understand such Tech Concepts and Machine Learning Case Studies. Download our Tech for Product Manager having Concepts and Case Studies about Artifical Intelligence