Okay, this is actually very clever: map out the world in 3x3m (roughly 10x10ft) squares. Algorithmically create a three-word code for each of those — figure.heat.museum or match.shutting.petted or holds.cheaply.fussed — and now, all of a sudden, people can get to anything.
This works great in areas that lack street addresses (a relatively new development in the world’s history, and something often missing even in certain areas of first world countries), or where the street address is misleading (“Meet me at the back door — takes.chimp.rice”) or imprecise (“We’re on the bridge at repaying.dodgy.statuses“) or non-existent (“We found a good place to picnic — trophy.mappings.communal“). And the words are easier to say and remember (and validate) than strings of latitude-longitude.
Apparently Mercedes is going to be incorporating What3Words coding in their voice-controlled satellite navigation system, and there are other companies looking at it for a variety of other uses. The biggest issue I see at the moment is that it is a proprietary setup — you have to license the algorithm/API from the company, so it’s not at all clear that you will be seeing it in Google Maps (or other mapping tools) any time soon.
But conceptually, it’s awfully cool. Maybe Google will buy them instead of the service.
what3words | Addressing the world
what3words is the simplest way to talk about location. We’ve divided the world into a grid of 3m x 3m squares and assigned each one a unique 3 word address.
More story here.
http://blog.caranddriver.com/just-three-words-can-get-drivers-exactly-where-they-want-to-go/?utm_source=feedburner&utm_medium=feed&utm_campaign=Feed%3A+caranddriver%2Fblog+%28Car+and+Driver%29
But who picks the three words? I mean, I can see battles of this. You might want the spot outside your house to be dave.house.road but what if somehow you get crackhouse.boondocks.dave instead?
It's been out there for a while.
https://www.kahomono.com/2016/03/three-word-coordinates/
Spinning of another conversation about this:
Other than ease of recitation (which we don't generally do in the GPS era), how is this better than existing lat/long coordinates – better enough that someone would want to buy a license?
+Michael Verona In the use cases I can think of, the verbalization of coordinates has a lot of applicability, and the words themselves provide a better checksum than a string of digits.
The specific application I saw being trumpeted here was verbally controlled satnav in new Mercedes, where, again, giving a set of words is (if the word recognition is decent) better than reciting a string of digits, esp. if conventional street addresses are not available.
+Dave Hill – The verbalizing for the satnav makes excellent sense!
+Laura Ess My assumption is that (a) the specific words are algorithmically generated based on the coordinates of the square, and (b) the vocabulary has been vetted to be relatively value neutral; this is assisted by a number of applications where the 10x10m is excessively precise, so that people looking for a w3w for their farmstead might have a few choices.
It's worth noting that adjoining squares have no verbal elements in common. That keeps it from being of use trailfinding ("I'm passing 2500 Main St, so 2600 should be coming right up") but also also serves as a bit of a checksum as well ("Wait, the map says colonial.golf.indigo is in New South Wales … did I get the wrong address?").