A new innovation has landed in the world of geography with the conception of what3words.
What3words (w3w) works by dividing the surface of the earth into a grid of squares “very close to” 3 meters by 3 meters in size (the nature of a globe is such that the surface cannot actually be divided into identically sized squares). Each grid square is represented by a 3-word-combination and the center point of each grid square is assigned a latitude and longitude pair. According to the technical report available on the what3words website, “w3w addresses are expected to be used in circumstances where an address, a verbal, or a written, description of a location would normally be used and, compared to each of those, they have the advantage of being more precise than any of them.” However the system is not designed for mapping or surveying purposes.
This looks to be useful as a world-wide geographic addressing solution, especially for areas that do not have well-defined, existing addressing systems. And, geographically, it would be an improvement over existing systems that are not designed to be used for determining geographic location but are, like the US Postal system which is intended for the purpose of delivering mail.
The words used in the combinations are “simply ordered words from a custom dictionary” made up of 40,000 words and they “have no meaning other than standing in for a large whole number.” Each word represents a number between 1 and 40,000, which, in combinations of 3, is plenty to represent the 56,666,666,666,667 3×3 meter squares it takes to cover the Earth. Words are used because words are easier for humans to remember than numbers, especially large numbers.
The service can be translated into other languages, but only English has enough words to cover land and water. So water addresses will always be in English, while land addresses can be translated.
W3w relies on a proprietary algorithm that converts the 3-word-combinations into a WGS-84 Latitude/Longitude pair representing the center of the approximately 3×3 meter square. The word combination is essentially a coded latitude/longitude pair. Which means that everyone who uses this system needs access to the w3w API in order decode the three word phrases.
Because the squares are so small, many 3×3 squares could refer to what is conventionally thought of as one location. For example, a single building may be covered by 80 squares and each of those squares has unrelated and seemingly random words assigned to them, which may create confusion if different people/databases don’t use the same exact square to refer to the location.
It will be fascinating to watch this system to see if it gains wide-spread adoption in the near future.