Cheltenham is a town of 116,000 people in south-west England. During the Covid-19 pandemic lockdowns, I spent lots of time collecting addresses in the town and adding them to the map.

Cheltenham is now better mapped than parts of London and many other cities and towns. If I can do it, you can do it too—and better still, you can learn from my experience and mistakes.

Why?

Address coverage in OpenStreetMap is still poor in many places. Good coverage is needed for better geocoding, which many users view as an integral part of a modern map.

During lockdown, many people relied on deliveries. Even now, a huge amount of time is wasted not so much in “last mile” delivery issues but “last few hundred metres” ones.

UK address data is not open, so addresses have to be surveyed or collected on the ground. This is tedious and repetitive, but there are some upsides. You’ll uncover mistakes in places which haven’t been touched on OpenStreetMap for years, and you’ll have the opportunity to improve geometry and alignment with the far better imagery now available.

Walking around your town or city is a great way of discovering new places and learning more about local history. Like many people, I’d heard of the mass trespass of Kinder Scout and how it ultimately led to greater access to the countryside. But I was unaware of earlier agitations here in Cheltenham over access to Leckhampton Hill. The hill is now part of the Cotswold Area of Outstanding Natural Beauty, crossed by the Cotswold Way National Trail which passes a memorial tree for one of the rioters.

Approach

I started collecting addresses in earnest in mid-April 2020, continuing to August 2021. I collected addresses at least once per week but usually more. I’d walk for anywhere from 30 minutes to four hours at a time.

Initially I focused on part of the town. By concentrating my efforts there, progress was easier to visualise, helping to keep me motivated. (Please don’t use a “scattergun” approach; go street-by-street, mapping comprehensively.)

There are various mobile editors such as Vespucci and StreetComplete (which now has an overlay for addresses), and the more recent Every Door (which didn’t exist when I collected addresses). I didn’t use any of them. Instead, I collected data and added it at home later as:

• Mobile editors are fiddly to use, especially in wet and cold weather.
• Often you’ll find buildings that are missing, need to be split, rotated, re-aligned, etc. You could use mobile editors only for simple cases, but then you’d have to make more complicated changes with other editors later. I find it easier to use one editor.

Experiment and use what tools work for you; consider what follows as a suggestion.

Tools

OsmAnd

I collected addresses using OsmAnd, recording a GPS trace as I walked, adding waypoints for addresses. This is basic but flexible; you can easily note other things to be changed or added, and you’ll likely develop your own system of abbreviations to become more efficient.

Here’s what OsmAnd looked like on my phone after an address collection survey:

I copied the GPX files from OsmAnd to my laptop and opened them in JOSM.

JOSM

JOSM is great! Yes, it’s unusual to use at first, and some things about it still irritate me, but it’s so powerful. Take the time to learn it and you shall be rewarded.

Here are just a few of the things I find useful in JOSM.

A superior validator

JOSM’s validator is powerful. It warns for things like:

• Duplicate house numbers on the same street
• House number without a street
• Similarly named ways (can indicate typos)

The validator caught several of my mistakes.

Coloured Streets

JOSM offers lots of different rendering styles. Coloured Streets is really useful for addresses. It makes it easy to spot certain kinds of mistakes. A friend described it as “syntax highlighting for streets and addresses”, which is an interesting way of looking at it.

See my Coloured Streets diary post for more details.

Buildings tool

You’ll likely find missing buildings, and the JOSM buildings tool is fantastic for quickly adding buildings to the map in two or three clicks. You can easily make buildings parallel to others or to roads.

See also Mapbox’s excellent fast building tracing tutorial.

Often the buildings tool and extrude are mentioned together. That’s fine, and although extrude is very useful for buildings, it can be used for much more. You can extrude a node along a way, which is great if you want to extend a cul-de-sac’s final node while keeping the road straight.

Terracer

Terracer is excellent for splitting large blocks of terraced houses (also called row houses or townhouses) into individual houses.

Only late into my address mapping did I discover a useful time-saver: if you’re splitting a building into two semi-detached houses, you don’t need to enter the highest number. Terracer determines it for you. It’s a small saving but adds up.

Angle snapping

Angle snapping lets you add lines with fixed angles between segments.

Imagine you’re adding a driveway which runs perpendicular from a road and then turns 90°. With angle snapping, when you add a node, the next segment of the way can be at a precise 90° (or other angle) to the previous. No more fiddling trying to make sure the angle is exactly what you want.

Circle arc

Circle arc is part of the useful UtilsPlugin2 and makes it easy to create proper curves and corners. I use it for circles too, as JOSM’s built-in “Align nodes in circle” feature often produces poor results.

UtilsPlugin2 includes some other essential features:

• “Split area”: split a closed way between two points. Useful for splitting a complex or irregularly-shaped building into two semi-detached houses.
• “Add nodes at intersections”: this is useful if you’re adding a service road that goes through a building. Select the road and the building, and points will be added where the road and the building meet. Then you can split the road and tag the section that goes through the building as covered=yes.

OSMUK cadastral parcels

During my address mapping, OSMUK created a public cadastral parcels overlay. You can show it on top of aerial imagery to see land parcel boundaries. It’s useful for figuring out how many houses are in a terrace, and also useful for alignment.

It’s available in iD and JOSM.

Overpass Turbo

Overpass Turbo allows you to query OSM data and visualise the results on a map. I used it for seeing which buildings were missing addresses:

[out:json][timeout:60][bbox:{{bbox}}];
// gather results
(
  way["building"]["addr:housename" !~ "."]
  ["addr:housenumber" !~ "."]["building" !~ "construction"];
);
out qt center meta;

… and also for visualising the addresses that I’d added:

[out:json][timeout:60][bbox:{{bbox}}];
// gather results
(
  nw["addr:housenumber"];
  nw["addr:housename"];
 );
nw._(newer:"2020-01-01T00:00:00Z")(user: "nickjohnston");
// print results
out center;

Leigh Dodds wrote an excellent Overpass QL tutorial if you want to learn more.

If Overpass Turbo seems like too much work, try Simon Poole’s QA map (tick “no address” or “has address”).

Strange addresses

You’ll soon discover plenty of strange addresses. My favourites are:

• A house with an address on a street that no longer exists (56 Worcester Street)
• Two houses numbered 10 on the same street (Upper Bath Street)

Advice

So you want to start collecting addresses? Great. Here’s some practical advice for surveys:

• Wear a high-visibility vest. Having people think you’re a delivery driver or parking warden is a lot better than a burglar.
• Bring a power bank and charging cable. If buying a new power bank, make sure it supports USB PD. Even if your current phone doesn’t support USB PD, it allows the power bank itself to be charged much more quickly.
• Keep your power bank inside a dry bag.
• Carry plenty of water and snacks.
• Don’t worry about people questioning you. There are far more friendly and interesting people than nasty ones. I’ve had many interesting conversations with people. A café owner even gave me a doughnut for free!

Some thoughts on attribution

Attribution can be a controversial topic within the OpenStreetMap community. Many people believe in clear, on-map attribution for maps based on OpenStreetMap data. Others are happy for attribution to be shown only when tapping an “i” or similar icon.

Now that I’ve spent hour upon hour out in rain, sun, snow, wind, heat, and cold collecting data (and later adding it to the map), I better understand why people feel so strongly about this. I’ve used OsmAnd so much that static parts of its UI are burned into my phone’s OLED screen. I’ve had a bird poo on me. Someone called the police who stopped and questioned me. So I’d say it’s not too much to ask that people attribute OpenStreetMap clearly and directly.

Visualisation of my progress

Pro-tip: with this tiny bit of MapCSS, Overpass Turbo results for buildings without an address are much clearer (small markers):

{{style:
node, way, relation {
  symbol-size: 1;
}
}}

Some thanks

Huge thanks to people in the #osm-gb IRC channel for their help, advice, and encouragement.

Also thanks to a friend (you know who you are) who twice gave me snacks and cold drinks when I was surveying near his house.

A section of the Cotswold Way National Trail on Leckhampton Hill was recently improved to make it suitable for off-road Tramper wheelchairs and mobility scooters.

The path has been widened and resurfaced:

Steps where the path meets Hartley Lane have been replaced with a ramp:

I’m not sure how best to tag this. Using wheelchair=yes doesn’t seem correct as a more rugged chair or mobility scooter is required. Looking at Taginfo for wheelchair I don’t see “off-road” or similar. The Key:wheelchair wiki page doesn’t help either.

Even the experts on the #osm-gb IRC channel seemed unsure how to tag this, but pointed out that such paths are becoming more common (see Rambles on wheels), so I thought I’d ask the wider community.

Better accessibility tagging is also useful for people who don’t use wheelchairs, but have restricted mobility. Someone at the volunteer group which helps look after the hill told me that the ramp has opened up the hill to people who have been unable to walk there for years.

For months I’ve been mapping addresses in Cheltenham, England. I recently discovered the JOSM “Coloured Streets” style. It has improved my address mapping and might help you too.

The style shows streets and all the buildings and address nodes “on” them in the same colour. This makes it easy to spot buildings tagged with the wrong addr:street at a glance. It also clearly shows buildings with addr:housenumber or addr:housename (you must enable the latter in the settings) but no addr:street. (The JOSM validator checks for this too, but with Coloured Streets you can see the problem immediately.)

In this example, Ashcombe House, Outwoods, and Beechwood are in pink and therefore have an addr:street of Stanley Road, when they should be on Ashley Road. The mistake would not be obvious otherwise.

If you look near the bottom right of this screenshot, you can see Windsor Street shown in orange:

Going north, the street becomes Cleevemount Close–but something is wrong: buildings continue in orange (meaning they have an addr:street of Windsor Street) alongside Cleevemount Close, which is purple. The house numbers on this section of the street are consistent with the numbers on Windsor Street. The street alongside 58-68 is probably still Windsor Street, not Cleevemount Close. I double checked on the ground, and the street is indeed Windsor Street. I’ve fixed the problem.

Coloured Streets makes problems like these more obvious.

You can activate the style in JOSM by:

1. Selecting the “View” menu.
2. Selecting “Map Paint Styles”.
3. Selecting “Map Paint Preferences”.
4. Selecting “Coloured Streets” in the “Available styles” list and pressing the arrow to add it to “Active styles”.

As with any other JOSM style, it’s easy to toggle on and off, or to temporarily hide by switching to wireframe mode if the colours become overwhelming or distracting.

The need for addresses

One of the things that often comes up in discussions about OpenStreetMap is poor address coverage in some places.

Rightly or wrongly, to many people a map is now no longer just a 2D representation of our environment. Searching and navigation/routing, despite being separate services, are viewed as part of a map, and almost as important as the visual map itself. We need good address coverage to improve geocoding.

Adding addresses really improves the map, and that’s what keeps me going with it. When I explained what I was doing to one person who approached me, he said “that’s a good idea mate, Google Maps shows my house all the way down the end of the road”. I also saw a food delivery driver pull up outside a house and ask “is it called Gate House?”. (On OpenStreetMap it is ☺)

Chancel Way or Chancel Park?

Recently I collected addresses on Chancel Way in Charlton Kings, Cheltenham. When adding them, I noticed something that didn’t seem right.

The road was split into Chancel Way and Chancel Park. At first glance, this seems fine. Having a common prefix like this is common in UK residential street names. (For example, Croft Road, Croft Gardens, Croft Drive, and Croft Avenue are nearby.)

The house numbering suggested that Chancel Way and Chancel Park were actually a single street. Following the street west, the first number on the left (south) side of Chancel Park was 19, and the next number was 21. This fits with the final odd number on Chancel Way (17). Even numbers on the right (north) side of the street match too.

More research

There are no results for Chancel Park on the Royal Mail postcode finder, yet it appears on Apple Maps (via DuckDuckGo):

It also appears on TomTom (no surprise as Apple uses TomTom data in places):

The mysterious Chancel Park also appears on Bing Maps:

… and HERE.

Much as it pains me to admit, Chancel Park is not shown on Google Maps; the whole street is shown as Chancel Way.

I even walked back and checked for any sign of Chancel Park on the ground. I found none. Chancel Park does not exist.

How did the mistake happen?

Chancel Park was added in changeset 7794450 on 2011-04-07, over nine years ago. The source:name tag was set to OS_OpenData_Locator. To my surprise, OS OpenData Street View shows Chancel Park:

OS MasterMap correctly shows the whole road as Chancel Way though.

It’s likely that TomTom (who supplies Apple) and HERE trusted OS OpenData Street View, and ended up with the mistake that way.

Closing thoughts

I discovered this mistake (which I’ve now fixed) through address mapping. If I’d just looked at the area on OpenStreetMap, the problem wouldn’t be obvious. Even a normal ground survey wouldn’t pick this up unless you were specifically comparing streets on the map with streets on the ground.

While out collecting addresses I’ve also found other useful things to map including:

• Tiny missing paths and shortcuts (like this one)
• Missing turn restrictions
• Cycle lanes
• …

Looking at other maps for preparing this entry has made me realise how plain they are and how little information they convey compared to OpenStreetMap. Apple Maps has no buildings in this area, and doesn’t show River Chelt either.

Most striking though is how little information other maps show for walking and cycling. There’s a path linking St Judes Walk to a walking and cycle path to the west. This is a pleasant route through Charlton Park and Cox’s Meadow, and then on to the hospital, lido, Sandford Park, and the town centre. This level of detail is a huge advantage of OpenStreetMap.