When Google Maps turned 15 on February 8, 2020, Senior VP Jen Fitzpatrick acknowledged what a gargantuan, never-ending task it is to create a truly helpful map. There are no licensing fees or other costs. The biggest advantage of Plus Codes is that unlike other alternate addressing systems available today, they are open-source and free-to-use. For example, in Plus Code parlance, Google’s HQ can be described as either “849VCWC8+R9” or “CWC8+R9 Mountain View”. The code can also be combined with a locality. Each code is 10 characters long and consists of two parts: an area code (roughly 100 x 100 kilometers) and a local code (about 14 x 14 meters). After all, the World Bank estimates that half of the world’s urban population lives on an unnamed street.įor this, Google Maps and Google Maps Platform support Plus Codes.Ī Plus Code is an address based on the latitude and longitude coordinates of a location. But what happens when a building does not have a traditional address. Today, Google Maps contain information, ratings, and reviews of about 200 million places around the world. This means, in essence, Google Maps is updated constantly – literally, every second of every day! That’s more than 200 contributions every second about things like road closures, local store details, the opening of a new business, etc. Though any new piece of information is updated only after ensuring a high degree of confidence in its accuracy, Google Maps receives more than 20 million inputs from its users every day. Google’s passionate community of Local Guides, active Google Maps users, or even the business owners that use Google My Business services, are continually providing new information to Google. This is where real people enter the picture. And these components alone do not always give users the context they need about a specific place on the map. Google Maps imagery, especially, can be anywhere from 1 to 3 years old (even older, in some cases). Improved coverage of buildings (pink) and points of interest (green) in Lagos, Nigeria from 2012 to 2018Īlso see: How Google Maps is using machine learning to ease our parking woes By using machine learning-based approaches, Google is now able to map as many buildings in one year as it previously used to map in 10 years! In Lagos, Nigeria alone, machine learning has helped Google add 20,000 street names, 50,000 addresses, and 100,000 new businesses. To manage the vast amounts of incoming data, machine learning libraries and frameworks are used. Overhead imagery gives Google details of roads and buildings, while street-level imagery provides road names, road signs, building numbers, and business names. Imagery: The imagery comes from broadly two sources – aerial images captured by airplanes and earth-observing satellites, and street-level data obtained by Google’s own Street View project. Think United States Geological Survey (USGS), Mexico’s National Institute of Statistics and Geography (INEGI), local municipalities, housing developers, etc. Well, the basic thing to understand is that there are two key components to mapmaking: data and imagery.ĭata: Google Maps use data from more than 1,000 authoritative sources around the world. With so many critical applications depending on Google Maps for accuracy and seamlessness, wondering how frequently does Google update its Maps data is quite reasonable. People rely on Google Maps and Google Maps Platform to not just get from point A to B, but to run businesses, to order food, to hail a cab, or to even provide SOS alerts during emergencies. And every week, more than 5 million active apps and websites leverage the core products of Google Maps Platform. Today, more than a billion people use Google Maps every month. In 2005, Google launched an ambitious project to map the world.