The Secrets Behind Google Maps
Billions of people regularly use Google maps, a tool that generates $11 billion in revenue, yet few know how it was developed and how it works.
Over time, rumours have spread that Google Maps was developed by continuously tracking the location and travel data of its users without their consent, allowing Google to create an almost perfect representation of the known world.
The Acquisition Spree Behind Google Maps
Google Maps is an online mapping and navigation platform used by one billion people monthly, as a navigation tool, providing directions as they travel from one location to another.
Google Maps is owned and operated by Google. Like many other Google products, however, the original mapping software was not developed in-house.
The navigation platform most have now become accustomed to started as the brainchild of three startups: Where 2 Technologies, Keyhole, and ZipDash. The contributions of Where 2 Technologies, however, were the most influential.
In 2004, Google acquired Where 2 Technologies, along with ZipDash, a company specializing in real-time traffic analysis, and Keyhole, a geospatial data visualization company known for capturing satellite images of various locations.
However, Google’s commitment to enhancing Google Maps didn’t end there. Since 2004, the tech giant has invested over $2.5 billion in acquiring various companies, each contributing unique technologies to the continuous evolution of Google Maps. Notable additions to Google’s portfolio include Endoxon, ImageAmerica, Quiksee, Zagat, Clever Sense, Waze, Skybox Imaging, Urban Engines, Sigmoid Labs, Pointy, and Breezometer.
The Art and Science of Maps
The main features of Google Maps are route navigation and real-time traffic analysis.
With route navigation, Google Maps can show users an accurate path from any point A to B of the developed world, and with traffic analysis, Google Maps can show users whether or not there is traffic on a particular route.
But how do these work?
Route Navigation
Whether you are going from your home to a friend’s place, from Spain to Singapore, or from the Liwa desert to civilization, Google’s Route Planner shows users possible routes from point A to point B.
But how does this work? How has the team at Google created software that can show directions to 98% of occupied places on Earth?
To achieve this, Google employs a six-step process: Data → More Data → AI Processing → Even More Data → further AI Processing → and lastly, Human Validation.
Above everything, Google is a data company, and it relies heavily on collecting, processing, and presenting a vast amount of data in its product offerings.
Fun fact: Google stores about 15 exabytes of data on its servers worldwide, which is equivalent to the storage space of about 117 million base model iPhone 13s.
Firstly, the Google team starts by getting “Data”. Google partners with thousands of public and private organizations around the world to collect domestic and geographical data such as city layouts, road plans, aerial imagery, business addresses, transit routes and schedules, etc.
The information from these organizations, however, just serves as a base model upon which the maps are built, as they could be incomplete or outdated, and that’s where “More Data” comes in.
So, to fill the data gaps from the organizations, the Google Maps team collects its own data, by deploying scores of Street View cars to the areas of interest.
Street View cars: These are vehicles with advanced cameras mounted on them to capture 360° street-level images of different locations.
The Street View cars drive slowly along the streets of the target area, capturing thousands of images of its surroundings. During this process, the Street View cameras take images of road signs, buildings, intersections, traffic lights, offices, gas stations, restaurants, etc.
Then, all the captured images are passed through Google’s proprietary AI technology which compares previously collected mapping data to the new ones, validating them and making changes where necessary.
For instance, suppose the foundational map data indicates that Sally’s Ice-cream shop is situated on Birch Street. The AI will examine the images captured by Street View cars along Birch Street to verify its current existence. If the shop is identified, the AI affirms its presence; conversely, if the AI cannot locate the shop, it will remove the outdated information from the map.
But still, this level of data isn’t enough. The Google Maps team aims to create a near-perfect navigation system, and for that, “Even More Data” is required.
Google Maps continuously relies on satellites to capture top-down images of houses, buildings, and streets, to further verify and modify previously collected data.
The satellites capture minute details such as the zebra crossings, the width of streets, shapes and dimensions of buildings, etc. All this information is then superimposed with the previous map data to create an accurate 2D view of cities. See the example below:
Now, the Maps team is satisfied with the data it has collected, and all this information is combined, processed, and incorporated into Google Maps by Google Maps’ AI and Machine Learning algorithms.
Lastly, human employees at Google make adjustments based on feedback from the public.
Note: In developing countries, like Nigeria, where mapping data from local organizations is unreliable and Street View cars can’t be deployed, Google Maps relies solely on images from volunteers and satellites to map out different locations, which leads to fewer mapped-out areas and poor navigational accuracy.
Real-time traffic updates
Another interesting feature of Google Maps is that it lets users know what routes have the least amount of traffic, saving users time in their commute.
But how does this work?
A few years ago, when Google Maps first introduced this feature, it relied on cameras mounted on traffic lights at various intersections to estimate the traffic conditions of specific areas. This method, though, proved to be inefficient.
Google Maps now collects the location and speed data from millions of smartphones with the Google Maps app installed and uses this data to estimate the traffic conditions on different roads.
Further Explanation: Google Maps looks at how fast everyone is driving on different routes. If Route A has a lot of slow-moving phones and Route B has a lot of fast-moving phones, then Google Maps suggests Route B, as there is less traffic on that route.
The Future of Navigation
Google Maps has become an indispensable part of our lives, providing individuals and businesses with accurate maps and navigation across the globe.
As Google Maps continues to push the boundaries, the future of navigation is immersive and intuitive thanks to AI and augmented reality.
With revolutionary AI capabilities, such as neural radiance fields, users can be shown realistic 3D digital models of cities around the world, allowing them to experience the vibe of a place before even visiting.
Also, augmented reality features will further enhance indoor navigation by providing heads-up guidance within intricate spaces like airports and malls.
The evolution of Google Maps since its inception is nothing short of remarkable. From the humble vision of four Australian entrepreneurs, it has grown into a service that not only maps out the entire world but also shapes the way we navigate and experience our surroundings.
Comments
Post a Comment