Data drop – Bus Stop Localities

London has continuously grown over the centuries, with conurbations and smaller dwellings overlapping and joining together to form the greater London area that we know today. This means that the ways in which areas of London are defined can often be inconsistent or confusing to those not familiar with the city, and we aim to enrich our bus stop location data to offer a wider geographical context and to help people understand London’s localities at a glance.

(Left) Borough, Postcodes, Stations, Wards. (Right) An example of perception of local areas

What would you call the area you work or live in?

When it comes to navigating around the capital and identifying areas – the data that currently exists within organisations like TfL, Greater London Authority (GLA), local authorities, central government, Royal Mail and Ordnance Survey (OS) varies from the wider perspective of borough boundaries through to postcodes and electoral wards down to the locations of stations, like on the tube map.

Yet these definitions don’t specifically describe an area at a more local level. A borough covers too wide an area – a postcode / postcode sector is setup for the business rules of delivering mail and have no name, the ward boundaries are smaller and more accurate but have more obscure and less well known names, and a station is just a specific point.

The concept of named places

From a wayfinding perspective it is sometimes easier to ask people for directions to an area that is well known, rather than a specific point or road that is lesser known. This is a concept Legible London has been championing as part of their guiding principles – for all users to process information more easily and understand the layout of London by creating a role hierarchy of named places; from ‘Areas’ at the broadest through to ‘Villages’ and ‘Neighbourhoods’ at the more precise end.

The Legible London, ‘The Yellow Book: a prototype wayfinding system for London’, discusses this in much more detail.

So what is a good method for defining a place boundary?

The short answer is that there is no precise way to define a named place – it is all speculative.

Using the centroid of an area / place and a radius throws up a number of questions – What is the centre? Is there more than one centre? What about overlaps with other areas? Do people prefer to associate themselves with an area of a specific socio-economical background? etc.

A good example is Chislehurst. Firstly, is it in south east London or is it in Kent? Secondly, you could say the centre is the High Street or the cross-roads at the war memorial, or the main train station which is a good kilometre away, and then it starts to overlap with Elmstead and Bickley.

However, the GLA have worked on developing a definition of named places ‘localities’ which closely maps to the Legible London concept of ‘Villages’. This dataset was inspired by a map of London’s ‘communities and open space’ in Abercrombie’s 1943 County of London Plan, and is based on information from a range of sources; these include Ordnance Survey maps and historic growth patterns.

The aim of this is to describe London in a way that is more recognisable to Londoners than Ward boundaries or statistical areas.

Bus stops and wayfinding

Bus stops are a great place-marker, and there are many of them across the Greater London area, but they currently lack any wider geographical context as to their location.

We have therefore trialled a one-off exercise of enriching our bus stop location data with the GLA’s locality data to create a ‘suggested locality’ field for the majority of bus stops. We are using the term ‘suggested’ as opinion at a local level can be very subjective, and there will never be a definitive consensus.

The dataset is not intended to be a definitive record of London’s socio-geographical structure, but an aid to help understand London’s localities at a glance.

Bus stops grouped by a perception of locality
Bus stops grouped by a perception of locality – the ‘suggested locality’ field.

The dataset is available on the documentation page of the developers portal under the heading ‘Bus Stop Localities.’

Our aim is to develop the places and boundaries dataset with the GLA to closely reflect a London which is recognisable by many, and which will benefit both wayfinding digitally and signage locally. We need your help to do this.

Feedback on this dataset

We envisage developers using this data to help their app users – our customers – to orientate themselves to their surroundings at a more local and meaningful level. We’d like your opinion on the definition of place names and the suggested localities we have given to specific bus stops – do you agree with this method of ‘suggested localities’?

We would also welcome any comments on the usefulness of the data set and other ideas or potential uses. Please do leave a comment in the comments section below.


  1. I’m not sure that customers will associate some of the stops with the localities they’ve been assigned – especially in central London where some of the boundaries are arbitrary – different people will associate places in Waterloo, Borough, Southwark, or Blackfriars respectively. In that way they may be given information which doesnt feel sensible?

    A different approach might be to do cluster analysis on the search terms being used in journey planner apps? Identify which locations are non-geographically specific (not postcode or station) and then investigate whether there are alternative options which might be giving them the wrong location? For example if a high number of people are searching for Colombia Road Flower Market but we don’t recognise that as a returnable location? This might produce more organic boundaries between places and a geography that reflects customer travel patterns?

    Alternatively we could examine the travel catchment of these search terms and complete a hierarchy based on the distance people are trying to travel to reach these destinations and use those to define the size of the locality, i.e. the attractiveness of the place to visitors? So if a large number of people are travelling from outer London to Borough market then you define that as proportionally larger to encourage people, if they use the bus, to use a wider variety of bus stops and prevent overcrowding?

    1. Thanks James, a couple of good ideas there we can use to develop the locality data set. Travel behaviour is definitely a high consideration.

Comments are closed.