Human and Machine-friendly: How Better Adressing Can Make Our Journeys Smoother

– by Clare Jones

With the growing trend for voice assistants – inside and outside of the car – it’s time we rethought a key part of that journey – how we input destinations. Street addresses weren’t designed for voice, and they lead to frustrating experiences in the car. Street addresses were developed to help deliver mail in a time when paper maps were the norm if they existed at all. They weren’t designed with voice input in mind; they were designed for buildings, not for entrances, places in the park for a yoga class, or the starting point for a hike. Street addresses alone aren’t fit for cars and journeys of the future, whether autonomous or not – they’re not accurate enough, they’re not universal and they weren’t built for voice.

ADDRESSES ARE SLOWING US DOWN
We’ve all had problems, at one point or another, with addresses and maps. Pins drop in the center of buildings with no indication of where the actual entrance is. Postcodes cover large areas. Duplicate street names cause mistakes. Satnavs tell us we’ve arrived when we haven’t really. New builds don’t appear on maps even though people have moved in and are using their new addresses every day.

Then there are all the places that have no address like food trucks, beaches, and fields, and all the developing countries and rural areas around the world that lack a formalized addressing system. Entering long street addresses, city names, and postcodes into satnavs is unwieldy and frustrating. Drivers around the world complain about the arrows and dials they are forced to use, and even if they have an on-screen keyboard, it’s an uncomfortable and typo-heavy experience. Addresses are also particularly difficult to enter correctly by voice. Numbers like 15 and 50 sound alike – and homophones are everywhere: no, I didn’t say “Lawn Road, I said Lorne Road”. Road names can be hard to pronounce for non-locals – Leicester Square or Worcester (pronounced lester and wooster) in the UK for example and speech recognition systems have to account for different pronunciations and accents in every country around the world.

A FUTURE-PROOF ADDRESSING SYSTEM
The first addressing system designed for voice, what3words has divided the world into 3m squares and given each one a unique 3-word address. For example, ///colder.episode.curvy takes you to one of the entrances to Arena hosting the STARTUP AUTOBAHN program in Stuttgart. 3-word addresses can be discovered on the free what3words app or online map, shared with others and navigated to, just like traditional addresses – only much easier to input by voice, and much more accurate than any street address. Guide books, hotels, restaurants and other businesses around the world are including their 3-word addresses on their contact pages so people can find them easily. With 3 dictionary words, people can input a precise destination, anywhere in the world. When entered into a navigation system or ride-hailing app, a 3-word address is converted to GPS coordinates in the background, simply providing a pin on the map to which the driver can navigate.

what3words was designed with voice in mind: each address is unique, homophones have been removed, and the AutoSuggest feature (where similar sounding words are far apart – ///toffee.branched.pyramid is in the UK, while ///coffee.branched.pyramid is in India) identifies and corrects any mistakes; what3words provides a 135% increase in address recognition vs. traditional street address entry by voice.

IMPROVING USER EXPERIENCE
The ability to tell a car, driver and friends exactly where to go removes the need for long phone calls asking for directions and reduces the distance traveled and time wasted. This means people have a better experience using a product, whether it’s a car, a delivery fleet van, a taxi, or an app that helps them get places. The system has been integrated into Mercedes-Benz’s new MBUX infotainment system and is now available in millions of cars in which drivers can now say ‘Hey Mercedes, take me to //grab.venue.glass’ to get directions to, in this case, Tabac Bar in London. Cabify, a leading ride-hailing app in Span-ish and Portuguese-speaking countries, allows users to enter 3-word addresses as destinations, so they get dropped off exactly where they want to. KakaoMap, the main digital map provider in the Republic of Korea has also included 3-word addresses in its map natively, so people can discover, share and navigate to them in the app they already use every day. Passengers can say 3 words to give Olli, an electric, autonomous bus, their exact drop-off point, and, speaking of electric vehicles, thousands of EV charging spots are now listed with 3-word addresses thanks to charging station-finding apps Moovility and evway.

IMPROVING USER EXPERIENCE
The autonomous vehicles of the future will need a much more accurate system for specifying destinations – telling a machine “oh the pin is always in the wrong place – keep going, it’s around the corner, the house with the red door” just isn’t possible.

If people are going to trust that their autonomous vehicles will take them to the right place, we need a radical change in how addresses are communicated between humans and machines. With what3words, people can ask their voice assistant to order them an autonomous taxi to ///left.clown.pasta, safe in the knowledge that, for a change, they’ll end up in exactly the right place.