KWANDENGEZI is a beguiling neighbourhood on the outskirts of Durban. Its ramshackle dwellings are spread over rolling green hills, with dirt roads winding in between. Nothing much to put it on the map. Until last year, that is, when weird signs started sprouting, nailed to doors, stapled to fences or staked in front of houses. Each consisted of three seemingly random words. Cutaway.jazz.wording said one; tokens.painted.enacted read another.
In a neighbourhood where houses have no numbers and the dirt roads no names, these signs are the fastest way for ambulances to locate women going into labour and who need ferrying to the nearest hospital. The hope is that signs like this will save lives and be adopted elsewhere. For the residents of KwaNdengezi in South Africa aren’t alone – recent estimates suggest that only 80 or so countries worldwide have an up-to-date addressing system. And even where one exists, it isn’t always working as well as it could.
Advertisement
Poor addresses aren’t simply confusing: they frustrate businesses and can shave millions of dollars off economic output. That’s why there’s a growing feeling that we need to reinvent the address – and those makeshift three-word signs are just the beginning.
“Poor addresses frustrate businesses and can shave millions of dollars off economic output”
In itself, an address is a simple thing: its purpose is to unambiguously identify a point on Earth’s surface. But, it also forms a crucial part of the way societies are managed. Governments use lists of addresses to work out how many people they need to serve; without an address by your name, you can’t apply for a passport or register to vote. That means an effective addressing system must do more than simply mark out a patch of land – it must also reflect the infrastructure there. One address shouldn’t span multiple houses, for example; postcodes shouldn’t split an apartment in two.
The favoured solution is one that comes naturally to most of us: a house number, followed by a street, a town and a country. The name for that combination is the , defined by the Universal Postal Union (UPU), a UN agency that works to ensure the 320 billion letters and nearly 8 billion parcels we send every year get to where they need to go. One of the main advantages of S42 is that it is hierarchical: houses are on streets in districts within towns inside countries. That makes it easy to group neighbouring properties together – a crucial asset for governments looking to evacuate citizens, or utility companies managing a district’s power supply.
But the system has drawbacks, too. An S42 address doesn’t tell you the physical location of a building, for example – for that, you need a list that can match it with a geocode, a lengthy combination of latitude and longitude, to identify the right spot.
Geocodes aren’t foolproof either, though. Just ask Chris Sheldrick, who used to stage private gigs. That meant getting performers and equipment to venues on time. “My attitude was: just give the driver latitude and longitude coordinates. They can stick them in their satnav; all satnavs accept them, and that’ll bring them to within metres of me,” says Sheldrick. But all those numbers can be hard to remember. One day, a driver got one digit of the code wrong and ended up an hour south of Rome instead of an hour north.
Sheldrick was still fretting over the affair a few months later, when he had dinner with an old friend named Mohan Ganesalingam. “I put it to him,” says Sheldrick. “There needs to be a human solution where people don’t make errors.”
That got Ganesalingam thinking. If you divided Earth’s surface into squares 3 metres across, would there be enough words in the English language to label each of the resulting 57 trillion plots of land with a unique three-word phrase? Simple arithmetic – the cube root of 57 trillion – showed that 40,000 words would be enough: a fraction of the .
Five years later, Sheldrick and Ganesalingam had put the idea into practice. Their what3words software is free to the public, but licensed out to logistics firms. With offices in west London – otherwise known as index.home.raft – they have also raised £15 million in venture capital.
It might seem strange that what3words is in vogue when there are plenty of other competing ways of dividing up the globe. Some use , others . , created by the founder of the TomTom mapping and navigation service, has been using squares since 2001. But they all use unmemorable sequences of letters and numbers to label locations, whereas Sheldrick and Ganesalingam were desperate to create something easy to remember. They latched on to a 1951 study by two psychologists at the University of Miami, Florida, showing that people’s recall was far better for words than for letters and digits.
Whatever the reason for its success, what3words is becoming a routine part of life in some parts of the world. The postal services of Ivory Coast, Mongolia, Djibouti, Tonga, the Solomon Islands and Sint Maarten in the Caribbean now use it as their . Steve Coast, founder of the crowdsourced mapping service OpenStreetMap, credits their marketing. “That’s why w3w is interesting,” he wrote in a blog post. “They have truly great – some of the best I’ve seen – people who’re pushing this solution all over the place.”
One of those places is KwaNdengezi. Coenie Louw, who directs a charity called Gateway Health Initiative, coordinates a community-run hospital transport service in the area that is cheaper and more reliable than ambulances or taxis. The service’s big issue was locating patients, says Louw. “We tried all sorts of things, like triangulating from cellphone towers, which never worked because they can’t give you an exact location.”
Now Louw sends out fieldworkers with smartphones to teach residents their what3words addresses. In the event of an emergency, these can then be used to help guide the community ambulances. Although he is still piloting the scheme, Louw reckons the typical wait is now about 30 minutes – down from the previous average of 3 hours.
But for all what3words’s usefulness, it doesn’t do everything an address should. A single house could have multiple three-word addresses, for example, while flats on different floors of a building would be indistinguishable. What’s more, the addresses give no clue as to how close or far apart they are. Sheldrick and Ganesalingam admit that their invention is just a smart way of specifying locations. “We’re the solution to a specific problem; we want to do one thing,” says Sheldrick.
But their idea of truncating coordinates into a more user-friendly code could still be more broadly useful. “An address is the currency you use to enjoy your rights and responsibilities in society,” says Patricia Vivas at the UPU. So assigning people a code that can fulfil some of an address’s functions without waiting for the authorities to get round to it could be a boon for unaddressed communities.
In 2014, the UPU launched a competition to wheedle out the best examples of geocoding innovation. Although what3words made it to the second round, it didn’t win. that did was proposed by Serena Coetzee and Victoria Rautenbach at the University of Pretoria’s Centre for Geoinformation Science in South Africa. Their idea was meant to help aid workers during the response to an emergency, such as a disease outbreak.
“An address is the currency you use to exercise your rights in a society”
Coetzee and Rautenbach set out how free satellite imagery could be augmented with an editable layer to assign each house a coded address. Fieldworkers could then use that to tick off addresses and upload information about them, creating a database as they went.
Of course, there was nothing inherently new about their ideas. They were simply a cheap way to help create and maintain an addressing system where no authority exists to do so. The one major advantage their version offered, however, was its editability. If new houses spring up in a neighbourhood, for example, or if a hospital wants to assign its entrances distinct identifiers to save time in patient triage, it can take ages for the official databases to reflect the change – even in countries where reliable addresses exist. An open-source addressing system could be the way forward.
But try creating something similar in a developed country and you could find yourself up against powerful interests. Almost 10 years ago, Richard Pope and Harry Metcalfe developed a simple web-based application that would take any UK postcode and convert it into a latitude and longitude. They hosted it at , named after the postmaster general who, in 1959, introduced what would become the postcode system. It quickly spawned several websites that used the tool to help people do nifty things, such as call up a list of applications to construct or modify buildings in a given area. It was also “a bit of a provocation”, says Pope.
That’s because Royal Mail and the government-owned firm Ordnance Survey own the authoritative geocoded lists of postcodes and addresses in the UK. Royal Mail, now privatised, makes about £30 million a year by licensing the data to firms ranging from couriers to insurance brokers, says Peter Wells, head of policy at the non-profit Open Data Institute in London. That means Royal Mail has a clear interest in defending its intellectual property. “They sent us nastygrams telling us to cease and desist – so we did,” says Pope.
The irony is that an open database allowing users to convert postcodes to geocodes for free would be a catalyst for innovation. Start-ups in fields ranging from home delivery to insurance would stand to benefit. Wells also points to the potential of driverless cars to automatically update such a database when they spot errors.
The economy, too, would benefit. Denmark adopted an open addressing system in 2002, and this boosted its economy by about €15 million a year, according to a study by the country’s . If you were to crudely extrapolate that to the UK on the basis of population size, says Wells, it would work out as an annual windfall of £110 million.
What’s galling for people like Wells is that open addresses seem to be catching on in most developed countries – France, for instance, introduced its database last year – making the UK a conspicuous outlier. A further irony is that at one stage, Wells had to create an open address list for the UK, which he reckons could have been done in as little as two years. The project died principally because he could not find insurers willing to cover it against the risk of litigation.
There’s no denying that creating and maintaining an addressing system is a complex and time-consuming business. Even the little things, like those eye-catching signs popping up in KwaNdengezi, have to be made by someone. But as ever shinier alternatives to the traditional address come to light, maybe the most important thing to consider is not what they look like, but who controls them.
This article appeared in print under the headline “Where in the world?”
