At work¹, I started creating a real pedestrian routing that uses sidewalks rather than the middle of the road (unlike most pedestrian routing software nowadays). This is necessary for calculating safe routes to and from schools. In this case, it really matters which side of the road the children use and which crossing they take.

The OSM community mainly uses two methods of mapping sidewalks (and the choice between these two methods directly influences the mapping of crossings): One option is to attach them as tags to the road (for example sidewalk=both), and the other is to use separate geometry. In this post, I would like to show an example where mapping the geometry separately makes a huge difference.

 

A Dangerous Crossing

The image shows one of Germany’s most dangerous places for pedestrians, according to accident statistics ². This crossing is located near Rothe Erde in Aachen.

The following figure, taken from a screenshot of JOSM, shows the OSM data for this crossing.³

As you can see, the crossing has five arms, one of which (the south east arm) is split into two separate lanes. There is a pedestrian crossing with traffic lights at each arm and tags reveal that there are sidewalks on both sides of each arm.

 

The Question

Let’s take a closer look at the two crossings at the north and north-west. Can you guess the geometry of these two crossings? Are they parallel to the next street (A in the figure above) or perpendicular to the street they cross (B in the figure above)? What do you think? Please try to answer this question, before reading on…

 

The Answer

Looking at the aereal image reveals, that neither is correct: One crossing is parallel to the next street (with the node of the crossing being placed slightly off) and the other one is (almost) perpendicular to the street it crosses:

What I want to demonstrate with this example: It’s impossible to determine the geometry of the pedestrian crossings from OSM data with attached sidewalks. You need separately mapped sidewalks to be able to determine this geometry.

 

Why Does it Matter?

Take a look at the next figure: Cars travelling from south east to north and pedestrians crossing the arm going north get a green light at the same time. This is called “bedingt verträglich” in German, meaning “limited compatible”, and is a standard for traffic lights in Germany.

The cars turn at an angle of 135°. This is considered dangerous because cars can travel much faster than they could at, say, 90°. This angle can be derived from OSM data. However, the position and direction of pedestrians cannot be derived: A pedestrian at position A might not see an approaching car because it is coming from behind him, while a pedestrian placed at position B will have a better chance of spotting the car in time because, in this case, the car is arriving from the left.

There is one more important detail: The distance from one side of the street to the other side of the street differs in both cases: In case A, it is 16.7 meters; in case B, it is only 13.9 meters. A difference of 2.8 meters, or about three seconds less of exposure, could make the difference between a safe and unsafe crossing. Again, this can be known with separately mapped sidewalks (if kerbs are mapped), but is almost impossible to determine otherwise.

 

This was just one brief example of an application of OSM data where the geometry of pedestrian ways is important and where this information cannot be obtained if sidewalks are mapped as part of the highway. There are more, and perhaps I will write about them in an upcoming blog post.

 

---

1. Research project SAFER: The project aims to create a smart pedestrian router to help plan better school routes. In addition to distance, other criteria like safety (which cannot be measured by numbers) will be taken into account. ↩

2. In a 2023 study, I analysed all cluster points of pedestrian accidents in Germany between 2019 and 2022. With 8 accidents, two of which resulted in severe injuries, this crossing was ranked 10th. ↩

3. I removed some unrelated details. ↩