Beginner's Guide to OpenStreetMap

 

 

1. What is OpenStreetMap (OSM)?

OpenStreetMap is similar to other mapping services except for one MAJOR difference. The map, in its entirety, was contributed and continues to be contributed by everyday people like you.

 

This leads to a direct advantage that OSM has over everyone else. The data is free to use and does not include any technical or legal restrictions. We'll cover why this is so crucial later in this section.

 

a. How did OpenStreetMap get the public to map the whole world?

It's important to remember that OSM was not forced on anyone. Everyone volunteered their time and geographic knowledge to this project.

 

If you can imagine, OpenStreetMap appeared to be a daunting task at first. Users with GPS units went out and tracked everywhere they went, then uploaded their data to OpenStreetMap. Still, all their efforts hardly resembled a map even after thousands of users began contributing.

 

However with each year that passed, the OSM community continued to experience a steady growth, and the maps have now become a very respectable source of information.

 

b. Are you kidding?

Yes, and no. It's true that much of the data found on OpenStreetMap was contributed by individual users doing fieldwork with their GPS units. However, some countries received a significant amount of help through donated data. For example, in the United States, TIGER data released by the U.S. Census Bureau helped enormously.

 

c. Is OpenStreetMap completed?

No, the true value of OSM is that anyone can edit it at any time. If you know of a new subdivision being developed in your neighborhood, just add it to OpenStreetMap! Re-routing of highways being done? Be the first to make the change on OSM!

 

While adding new features and moving roads tend to be the most fun and rewarding, OSM also needs quite a bit of editing done to their existing data. For example, features need to be "tagged" properly. Tagging is the equivalent of describing the feature - what's the street name? The speed limit? Elevation? Basically, every feature should be tagged with as much information as possible.

 

Also, especially with TIGER data in the United States, many roads are "over-connected" or perhaps not even connected at all. At first glance, it's possible users may not notice any inaccuracy in the data. However upon further inspection and with the help of other tools, they may then realize that an interstate freeway is connected to a residential road when in fact, there is a bridge separating the two. Fixing these issues will help further improve routing with OpenStreetMap.

 

d. Why should I contribute to OpenStreetMap?

The reality is you'll never know if a small edit you make in rural America will ever get noticed. But OpenStreetMap is about becoming a part of a community and contributing to a good cause. These maps can later be used to foster creative and unexpected uses of OSM, or even be uploaded to your GPS (for free, of course).

 

Recently, crisis mappers have generated a huge buzz for OpenStreetMap. When a devastating earthquake hit Haiti in January 2010, volunteers flooded towards OSM to help digitize much of the missing data. Their amazingly quick response undoubtedly contributed to the relief efforts of Haiti. Collapsed bridges and buildings, refugee camps, and road blocks were all marked on OpenStreetMap and assisted emergency response teams.

 

A seemingly inconsequential edit you help make today can very well affect another person's life, whether it'd be tomorrow or several years down the road.

 

e. Great! How Do I Sign Up?

Visit http://www.openstreetmap.org/ to sign up for a new account!

 

 

2. Choosing the Right Editor

Now that you are hopefully signed up for OpenStreetMap, it's time to choose an editor that's right for you. There are several options available with each having its own unique advantages, so choose wisely based on your technical needs and comfort level.

 

a. Potlatch

Recommended for beginners, Potlatch is a basic Flash-based editor. Because this is an online editor, no downloads or set-up is required. It's what you will find on http://www.openstreetmap.org/ when you click the Edit tab, and a newer version, Potlatch 2, is what is available through the MapQuest website.

 

Aerial imagery is already provided as a background to help with tracing and editing, and we also provide an extensive Potlatch 2 Primer to help beginners learn all the editing tools. An issue that some users may have, however, is its rendering speed. Because all data needs to be loaded from the OpenStreetMap server, experienced users may feel a slight frustration when panning and zooming.

 

b. JOSM

Popular among more advanced users, JOSM is a desktop application that requires downloading and configuring a few plugins in order to set up properly. The interface may also be intimidating at first with sidebars and tools galore but once you familiarize yourself with this application, very few people turn back.

 

Unlike Potlatch, however, JOSM requires you to download a small geographic area from OpenStreetMap onto your computer. The advantages to this method are plentiful. The panning and zooming are fast and fluid, GPS files and photos stored on your computer can be used as reference, and the vast number of available plugins make JOSM even more powerful and customizable to your needs.

 

Here are a few videos to help you get started with JOSM:

  

c. Miscellaneous

In addition to the editors mentioned above, there are a number of other applications that are used fairly frequently. Merkaartor is still in its early stages of development but offers some unique features such as transparent displays of roads and other features. GIS professionals with experience using ArcGIS10 will be delighted to know a plugin exists for editing OSM data called ArcGIS Editor for OpenStreetMap. Since new developments are always occurring, we'll try our best to keep you updated on other products as they become available.

 

 

3. Basic Quality Assurance Tools for Editing

Once you've decided which OpenStreetMap Editor you'll be using, it's fair to say you'll want to jump right into editing. However, most beginners will zoom to an area and realize everything looks correct. Aside from smoothing a curved road, or realigning a park boundary, there may not be any obvious errors. This is where the following tools come in handy.

 

a. OpenStreetBugs

OpenStreetBugs is so basic, in fact, that you won't be doing any physical edits here. (I know, we apologize!)

 

Available at http://openstreetbugs.schokokeks.org/, OpenStreetBugs provides a simple, easy-to-use interface which allows anyone to report an error they found on OSM. No login is required. Instead, you simply click the map at the location of the bug and leave behind your name and comment detailing the error.

 

Eventually, you or someone else can further investigate the issue and hopefully resolve the error. When the bug is marked as fixed, the icon changes from a red circle with an X symbol to a blue circle with a checkmark.

 

b. Skobbler

Similar to OpenStreetBugs, Skobbler allows the user to report bugs in OSM, have conversations regarding the issue, and mark them as resolved.

 

You can also search for the bug by ID, Status, Type, and Content.

 

c. Keep Right

Available at http://keepright.ipax.at/, Keep Right is a wonderful tool that shows automatically detected errors on the map or in list form in various languages. When you hover over an error, a pop-up window will appear with details and a link to make edits in either JOSM or Potlatch.

 

Here are just a few examples of what Keep Right is able to detect:

 

  • non-closed areas (parks or any other feature area that do not close the polygon)
  • one-way roads that come to a dead end
  • two roads that appear close but do not form a proper junction
  • deprecated or missing tags
  • places of worship without religion
  • POIs without a name

 

As you can see from the list, Keep Right is an excellent source to find relatively minor errors that you might not otherwise notice. Feel free to zoom in on your neighborhood and start editing!

 

Full list of automated errors are available on the Keep Right Wiki.

 

There's also an excellent video tutorial of Keep Right available at http://www.youtube.com/watch?v=8MF4oJIHuQw which provides basic editing techniques that we'll go over in detail later in this guide.

 

 

4. Advanced Quality Assurance Tools for Editing

 

a. Duplicate Nodes Map

The Duplicate Nodes Map also uses a simple map interface to show all locations of duplicate nodes.

 

While there are arguably some legitimate duplicate nodes such as boundaries that run along highways, most are due to a poor import of the data. For example, TIGER data was originally split by county and in some geographic areas, roads act as the defining boundary between two counties. This caused the same road to be featured in both counties and have resulted in duplicate streets and nodes to be uploaded to OSM.

 

Most major errors have already been taken care of, but the Duplicate Nodes map will help display all other instances that should be looked into since they are usually a good indication of a problematic area.

 

Additional information on why duplicate nodes should be investigated and how to resolve them can be found below in the Challenges in the U.S. Data section.

 

b. Waychains TIGER Fixup

As you might expect, the Waychains TIGER Fixup generates a waychains report. In other words, this tool reveals all unbroken sequences of ways that are tagged highway=motorway.

 

A way is considered unbroken when they all face the same direction, have the same int_ref value (highway number), and the motorway pair (northbound/southbound or eastbound/westbound) run in a general parallel direction.

 

This is important because a broken waychain on a motorway can potentially equate to:

 

  • Motorways and on-ramps being tagged incorrectly
  • Motorways facing the wrong direction
  • Motorways represented by only a single way
  • Possible duplicate nodes

 

While this tool requires some investigation of the data, the errors reported are no less important. Motorways facing the wrong direction tend to cause major routing problems!

 

 

5. Challenges in the U.S. Data

It doesn't take long for a new user to hear or read somewhere that OpenStreetMap in Europe is much more advanced than OSM in the United States. It's true. The U.S. map based on TIGER data still requires a bit of work. Between wrong way streets, over-connected roads, missing subdivisions, incorrect road classifications, and missing layer information, there appears to be no end in sight!

 

In fact, TIGER data also includes a special tag tiger:reviewed=no. This is just one way for OpenStreetMap users to keep track of which roads have already been verified as correct. (Once you've reviewed a road, please go ahead and remove the tiger:reviewed=no tag.)

 

Below, we'll go over common errors found in TIGER and how to resolve them.

 

a. Over-Connected Roads

Over-connected roads refer to intersections where junctions exist in OpenStreetMap but do not exist in reality. Let's take a look at an example below. Should there be a junction connecting these roads?

The answer is no. The road going across (in yellow) lies above the divided highway. Hint: You can see the slight difference in color where the actual bridge over the highway begins and ends as well as the bridge's shadow.

 

However, if a junction did exist, this could mean that any routing algorithm could call for someone to make a turn from atop the bridge to the road below! We're pretty sure we don't want OSM to be held responsible for vehicular accidents so it's important to verify that no junction exists.

 

By highlighting the road while Editing, we'll be able to check if junction nodes exist. If you see something similar to the image below (two black outlined boxes), a junction exists and should be removed immediately.

In Potlatch, you would remove the junctions by highlighting the node and pressing "-" on your keyboard. Once these two junction nodes are removed, making a turn from the bridge will finally be disallowed. While the routing data has now been corrected, there's also another edit you should complete - separate the bridge as another feature.

 

Split the road at the beginning and end of the bridge by first adding the node (shift+click), then highlighting the node, and finally pressing X (split tool). Once you have your bridge, be sure to tag it as bridge=yes and you're done!

 

If you had trouble following this tutorial, refer back to the Keep Right video above or at http://www.youtube.com/watch?v=8MF4oJIHuQw. About one minute into the video, it walks you through an OSM editing example very similar to the one shown above.

 

Hopefully you were able to see why over-connected roads cause routing problems and will continue to look out for them and resolve them as needed.

 

b. Duplicate Nodes

As mentioned earlier in the Duplicate Nodes Map section, entire roads could have easily been uploaded twice near county boundaries. Most of these duplicate nodes have already been taken care of, so we'll instead be taking a look at some minor instances of these occurrences that are also proving to be quite problematic.

 

One of the most common examples of a duplicate node needing to be fixed is of two or more roads that appear connected, but are actually drawn as two individual disconnected roads. Similar to over-connected roads, duplicate nodes may look visually correct but routing cannot be properly performed. If these roads do not share a common node, it is the equivalent of being a dead end street and no vehicles will be able to pass through.

 

To the right, we zoomed in near a county boundary on the Duplicate Nodes Map.

 

The red plus signs indicate all locations where a duplicate node exists and we can see clearly that they are all hovering over the road and no other features.

 

There's a logical explanation for this pattern we're seeing here. Because TIGER was created separately by county, merging county data most likely caused many of the disconnected roads.

 

Highlighting a road confirms what we thought was the issue. These roads are all disconnected. They may share the same point but there are two nodes here, one sitting directly on top of the other.

To resolve this in Potlatch, highlight one of the duplicate nodes and press Delete. This will remove the one node. Potlatch will automatically assume you want to continue drawing your line so now move your cursor back to where the duplicate node used to exist (now just a single node). The cursor should then change from the normal Pen tool to a Pen tool with a small black circle as seen below. Shift+click at the node to merge the two together.

It's possible you receive an error message regarding unmatched tags. In the above example, the two streets were identical except for the tiger:county tag. We'll look into fixing this in a little bit, but let's first take a look at the above image again. The county boundary, highlighted with a purple line, is farther west than where the duplicates nodes were initially placed (also where the roads' county tags were split).

 

Assuming this county boundary is correct, we must split the road at the point where it crosses the boundary. This way, we can tag the two sides of the road with their proper county names.

 

Add a node at this intersection (shift+click), highlight the node, then press X to split the road. Highlight one side of the road, then change the tiger:county tag to its correct value. Do the same for the other half of the road. Verify that no other tags need to be changed and that's it! One down, many many more duplicate nodes to go!

 

 

6. Adding Your Map Data

While it's become increasingly difficult to find areas that haven't already been mapped, they do still exist. If there is no aerial imagery to assist you with adding the new features, some fieldwork using a GPS will be involved. Here are the basic steps that must be followed:

 

  • Collect data using a GPS
  • Remove excess data (i.e. a track showing where you've stopped for lunch)
  • Upload data to OpenStreetMap
  • Clean up data in OpenStreetMap
  • Tag data

 

To learn more about doing fieldwork and collecting data, please refer to the Beginner's Guide on the OpenStreetMap Wiki.

 

OpenStreetMap data is licensed under the Open Data Commons Open Database License (ODbL).