Documentation

At its core, mobi.mapr is an analysis showcasing accessibility, but does this by taking subjective factors into account. This can be explained using an example.

A person is at home and wants to go to the next bakery. This bakery is either 12 minutes away by foot, 5 minutes by bike, or 8 minutes with the bus. With a standard accessibility analysis, taking 5 minutes with the bike would be deemed the quickest mode. In mobi.mapr additional factors are included in the scoring.

The result is an adjusted time, which is explained in more detail below. This is calculated and weighted for all modes and activities, and finally compared to the average time. This index is graded, resulting in our SIGMA score on a scale from A to F.

At the end of every trip, there is an activity. Whether it’s grocery shopping, the commute to work, or getting an ice cream. This journey to the activity generates the traffic we know.

A location is the physical manifestation of an activity; for example, the location for the activity Bakery is “Café Bäckerei am Schloss”.

Defining an activity is no simple matter. Is a bakery a place where you can buy bread rolls, or is it more of a place where you can have a cup of coffee? The mobi.mapr uses the classification from OpenStreetMap. Other classifications, such as those used by Google Maps, Apple Maps, or Overture, have proven unsuitable for this purpose.

Since a person engages in more than one activity, a logical grouping of activities is needed. Such a grouping is called a category. For example, if an activity is “supermarket,” a category might be “daily needs.”

This generalization makes it possible to build on well-known studies, such as Mobility in Germany, by translating the abstract activities used in those studies into actual activities seen in the real world.

Each category consists of more than one activity, each of which may have a weighting. For the “daily needs” category, a supermarket is more important than a bakery, in this case.

In addition, the categories influence the weighting of the routes for each activity.

Weighting

Weighting

Up until now, we’ve spoken about the score being calculated from a starting point to a destination. In reality, however, it’s often important to have more than one destination. For example, it’s important to have several bakeries available in case one doesn’t meet quality standards or has inconvenient opening hours.

The model always evaluates three alternatives to the fastest option, which are weighted according to category. The weighting function is as follows:

e^{-lambda * (i-1)} with i = 1,2,3

This results in the following weightings, depending on lambda:

To get from A to B, you select a mode of transportation. In the simplest case, you can walk, but other modes are also possible, such as driving, biking, or taking public transportation.

To make it easier to use, mobi.mapr distinguishes between profiles and modes. A profile is a variant of a mode. Here’s a simple example: In mobi.mapr, “cycling” is a mode, while “road bike” or “cargo bike” would be profiles. Strictly speaking, it should therefore be called “mode-profile,” but the short form is used instead.

A profile involves many subjective factors. Examples include the maximum speed the bike can reach in different configurations, or the time it takes to park the bike—which takes longer with a cargo bike than with a trekking bike.

The quality of mobility also always depends on a person’s location. If they are in a rural area, they tend to engage in different activities than in the city, and the average travel time varies.

mobi.mapr uses the RegioStaR7 classes, which divide Germany into seven different spatial types.

Based on these RegioStaR classes, different reference values are used, and the categories for personas (see below) are organized differently.

For more information on regionalization, check out the blog post “How to Compare Mobility Quality Across Regions”

If you want to aggregate multiple categories, it is important to give them different weights. After all, not all routes lead to places where people go for daily essentials, just as not all routes lead to children’s playgrounds.

mobi.mapr therefore uses personas as a way to synthesize these values. Version 1.0 therefore also uses the personas defined in Mobility in Germany.

This makes it possible to determine, for each municipality, county, or state, which categories are relevant and to what extent.

For more information on personas, check out the blog post “Von der Personengruppe zur Persona“.

To determine mobility quality, routes are calculated. These routes have the locations of activities as their destinations but require a starting point. The smallest spatial unit used in Germany is the municipality. However, this unit is still far too large.

In mobi.mapr, the area is therefore divided into hexagons. The smallest hexagon has a edge length of 200 meters. It is therefore possible to represent mobility quality at the following levels:

• Hexagon
  • Municipality
    • County
      • State

Once the routes have been calculated, a value must be derived from them. This value is called the Score in mobi.mapr and is, for example, 13 minutes.

Since mobility is typically compared relative to others, it makes sense to put this value into context. This context is the average travel time by regional class in the respective municipality. So if a trip takes 25 minutes on average, this value is below average. An index is created. It ranges from 0 to a maximum of 2, where it is capped for technical reasons.

To evaluate and finalize the results, a grade is derived from the index. This is divided into school grades, but to avoid confusion with the index, it is denoted by letters.

The intermediate steps are labeled with + and -. This is based on an exponential function, meaning that lower indices are weighted more heavily than higher ones.

The evaluation is based on the following criteria. If the index is greater than 100%, it is rated as unsatisfactory (E). The difference between “Good” and “Satisfactory” is 50% of the index.

This results in the following function: h(x)=0.2511169*1.122178^(x). For each x, the grades can be assigned in half-step increments; thus, x=1 corresponds to A+.

To ensure consistent notation, the symbol Σ (SIGMA) can be used at the end. It stands for Score-Index-Grade-Mobility-Accessibility. A municipality could therefore have a Σ B+.