Online maps are like manners. One only notices them when they are missing. But it’s actually worth understanding how they are working and how much work went into them. Justin O’Beirne from San Francisco made me aware in his detailed blog of the art and effort that goes into maps.
I blame it on my lack of mindfulness that I hadn’t become more aware a few weeks ago, when I went to Düsseldorf in Germany. To figure out where I should book my hotel, I was searching on google Maps for downtown. And I did notice the lightly orange colored areas, that seemed to mark downtown areas. I also have noticed for some time the building structures that Google Maps also displays.
And those details are exactly what Justin O’Beirne details in his blog. It’s fascinating how he uncovers the different layers in the maps and comes to conclusions for self-driving and electric vehicles. And it becomes apparent how far ahead Google is to other map makers such as Apple, TomTom, Bing, Here, or Waze.
While all maps display the most basic features such as roads in comparable quality and level of detail, Google doesn’t stop there but has way more details. And that starts with buildings. On the image below you can see the level of detail that Google and Apple display. Not only can you see structures, and structures right along streets, but Google also displays structures off roads. Parks are also colored differently with Google Maps.
This level of detail seems to be available on Google Maps for a few months only. Other map makers do have buildings for some areas, but only Google has them for almost all cities and settlements. And that means also small villages with only a dozen inhabitants. Those maps include a a variety of data sources, that allow that richness of information. Beside map data Google has been using Google Street View-cars. Google covered 99 percent of U.S. roads and cities, after driving Street View-cars for more than seven years. From those images collected more information is extracted. For instance, with the help reCAPTCHA they extract house numbers and store names.
The building structures apparently are extracted from satellite pictures with the help of artificial intelligence. As a byproduct of satellite and Street-View-pictures maps are enriched with building and store information.
Further enrichment is done through address data from stores and restaurants that Google already collected for its search engine through crawling the web. As a byproduct Google is able to tag commercial areas or city areas of high interest in the maps.
And Google is able to do that because they own the data. Unlike Apple, who seems to license its map data from TomTom. Google can therefore created derivative information that for example Apple cannot, as TomTom’s license agreement prevents that.
And that brings me back to downtown Düsseldorf and the light orange coloring. A large number of stores and restaurants indicates shopping streets that are of high interest for visitors and pedestrians. And this is how people navigate in cities. In their 2011 master thesis Visualizing Mental Maps of San Francisco, Rachelle Annechino and Yo-Shang Cheng asked 22 San Francisco inhabitants, to draw a mental map of San Francisco. The study authors discovered, that people navigate the city by such ‘areas of interest’
And that was exactly what I did when searching for a hotel. I wanted to be as close to downtown as possible.
Autonomous and Electric Cars
This level of detail benefits other types of consumers as well. I am talking about autonomous and electric cars. An address may not always be good enough for picking up a passengers. Uber users know that sometimes they have to direct the driver to the precise pickup point. A self-driving car that knows to stop in front of a stairway or that the house entrance is actually around the corner of the address specified makes it easier and less frustrating to pick up a passenger.
This information is invaluable for electric vehicles as well. Knowing the status of a charging station, which entrance to take, how many plugs are available and what charging speeds, what charging cards are accepted and at what time the stations are operating will save electric vehicle owner from frustrating experiences.
As the newest feature Google Maps now incorporates 3D information. Satellite images include perspectives when moving the map. Here is a video that Google published in that regard.
All thosse facts explain why Google is far ahead of other map and navigation system makers. Converging so many technologies and efforts that generate a huge amount of data allow Google to increase the information density of its maps and gives them an incredible head start in the development of autonomous cars.
And still, some of those technologies are not yet even part of that convergence. Google also operates an elaborate simulator to develop its robotaxis and have them drive flawlessly in urban areas. Once we start seeing convergence of simulators with maps, what kind of services will we see then?
A more detailed analysis of mapping features can be found on Justin O’Breine’s blog.
This article has also been published in German.